remote sensing Article Estimation of Forage Biomass in Oat (Avena sativa) Using Agronomic Variables through UAV Multispectral Imaging Julio Urquizo 1, Dennis Ccopi 1, Kevin Ortega 1,* , Italo Castañeda 1 , Solanch Patricio 1, Jorge Passuni 2 , Deyanira Figueroa 1 , Lucia Enriquez 1, Zoila Ore 3 and Samuel Pizarro 4 1 Dirección de Desarrollo Tecnológico Agrario, Instituto Nacional de Innovación Agraria (INIA), Carretera Saños Grande—Hualahoyo Km 8 Santa Ana, Huancayo 12002, Peru; juliocesarub3@gmail.com (J.U.); dennisccopit@gmail.com (D.C.); mict11000@gmail.com (I.C.); solanch.patricio.r@gmail.com (S.P.); deyanirafigueroa66@gmail.com (D.F.); luciacep7@gmail.com (L.E.) 2 Programa Nacional de Pastos y Forrajes, Estación Experimental Agraria Santa Ana, Instituto Nacional de Innovación Agraria (INIA), Carretera Saños Grande—Hualahoyo Km 8 Santa Ana, Huancayo 12002, Peru; jpassuni@inia.gob.pe 3 Dirección de Desarrollo Tecnológico Agrario, Instituto Nacional de Innovación Agraria (INIA), Av. La Molina 1981, Lima 15024, Peru; zoilaoreaq@gmail.com 4 Dirección de Supervisión y Monitoreo en las Estaciones Experimentales Agrarias, Instituto Nacional de Innovación Agraria (INIA), Carretera Saños Grande—Hualahoyo Km 8 Santa Ana, Huancayo 12002, Peru; spizarro@uncp.edu.pe * Correspondence: kevinorqu@gmail.com; Tel.: +51-992656272 Abstract: Accurate and timely estimation of oat biomass is crucial for the development of sustainable and efficient agricultural practices. This research focused on estimating and predicting forage oat biomass using UAV and agronomic variables. A Matrice 300 equipped with a multispectral camera was used for 14 flights, capturing 21 spectral indices per flight. Concurrently, agronomic data were collected at six stages synchronized with UAV flights. Data analysis involved correlations and Principal Component Analysis (PCA) to identify significant variables. Predictive models for forage biomass were developed using various machine learning techniques: linear regression, Random Citation: Urquizo, J.; Ccopi, D.; Forests (RFs), Support Vector Machines (SVMs), and Neural Networks (NNs). The Random Forest Ortega, K.; Castañeda, I.; Patricio, S.; model showed the best performance, with a coefficient of determination R2 of 0.52 on the test set, Passuni, J.; Figueroa, D.; Enriquez, L.; followed by Support Vector Machines with an R2 of 0.50. Differences in root mean square error Ore, Z.; Pizarro, S. Estimation of (RMSE) and mean absolute error (MAE) among the models highlighted variations in prediction Forage Biomass in Oat (Avena sativa) accuracy. This study underscores the effectiveness of photogrammetry, UAV, and machine learning in Using Agronomic Variables through UAV Multispectral Imaging. Remote estimating forage biomass, demonstrating that the proposed approach can provide relatively accurate Sens. 2024, 16, 3720. https://doi.org/ estimations for this purpose. 10.3390/rs16193720 Keywords: germination rate; machine learning; remote sensing; photogrammetry; vegetation indices Academic Editor: Mobushir Riaz Khan Received: 11 July 2024 Revised: 27 August 2024 1. Introduction Accepted: 5 September 2024 Several countries are exploring and implementing various novel strategies to improve Published: 6 October 2024 the availability of range forage for livestock in order to ensure food security, economic well-being and social cohesion of people [1]. Nevertheless, the increase in food demand and environmental pressure on soils have highlighted the urgent need to adopt more sustainable Copyright: © 2024 by the authors. and efficient agricultural practices and technologies. These should focus on mitigating Licensee MDPI, Basel, Switzerland. negative impacts and ensuring the long-term sustained production and efficiency of these This article is an open access article pastures [2–4]. Forages, which include a variety of crops such as grasses, legumes, and distributed under the terms and other species used for green fodder, hay, and silage, constitute the main source of livestock conditions of the Creative Commons wealth and are fundamental to these industries [5,6]. Being essential for livestock feeding Attribution (CC BY) license (https:// by providing the necessary nutrients for growth and production, it is crucial to increase creativecommons.org/licenses/by/ the rate of genetic improvement and conservation of forages to maintain the industry’s 4.0/). competitiveness [7,8]. Remote Sens. 2024, 16, 3720. https://doi.org/10.3390/rs16193720 https://www.mdpi.com/journal/remotesensing Remote Sens. 2024, 16, 3720 2 of 21 The most common grasses used as forage include forage corn, oat, wheat, barley, and rye grass, which are valued for their energy content and their ability to produce large volumes of biomass even in dry lands [9,10]. Among forage grasses, Avena sativa stands out as a highly important temporary grass worldwide due to its remarkable adaptability to a wide range of altitudes and climates [11]. In the Peruvian Andes, oat grows at altitudes ranging from 2500 to 4000 m above sea level and shows exceptional adaptability and high nutritional quality [12]. Its use, either alone or in combination with other forage legumes, enriches the protein content of rangelands, increasing their value as a food resource for livestock [13,14]. Agronomic variables related to yield, seedling growth, individual plant height, and others are essential as they play a crucial role in understanding and monitoring crop health and productivity [15]. They are also used to guide management practices, such as fertilizer application, irrigation, and harvesting [16]. Although conventional evaluation methods involve the use of manual measurement techniques and equipment, the limitations of direct observation make the process inefficient, time-consuming and prone to error [17]. Additionally, limiting measurements to only a few plants may not provide an accurate assessment of the entire field [18]. The use of emerging technologies that involve unmanned aerial vehicles (UAVs) in precision agriculture offers unprecedented covariables with high spectral, spatial, and temporal resolution, which could be linked to agronomic variables, deriving vegetation height data and multi-angle observations. Multispectral sensors represent a promising alternative for crop measurement and monitoring within the framework of precision agriculture [19]. This methodology relies on the intensive collection of spatiotemporal data and images to optimize resource use and improve agricultural production [20]. It also provides instruments and a variety of digital models to calculate plant height and other agronomic characteristics, which are utilized by UAV remote sensing technology through its various sensors [21]. This offers benefits such as ease of operation, flexibility, adaptability, and reduced costs [22], leading to a notable increase in its use within the agricultural research community [23,24]. In the context of precision agriculture, it is essential to have accurate phenological information to estimate agronomic variables from aerial images obtained with UAVs [25]. The precision agriculture approach using UAVs needs to be complemented with on-the- ground measurements, has demonstrated significant correlations, and has been successfully applied to a variety of crops such as maize [16,26], wheat [27,28], barley [29,30], and grass- lands [31–33]. To assess morphological variables using unmanned aerial vehicles, indicators such as plant height [18], germination rate [34,35], emergence [36], and biomass [37–39], among others, are used. These characteristics are employed in traditional linear regression algorithms and empirical models to predict crop yield and biomass, combining crop spectra with these agronomic variables [38]. These characteristics are used in traditional linear regression algorithms and empirical models to predict crop yield and biomass, combining crop spectra with these agronomic variables. It is also necessary to consider that these methods are already being applied in various vegetable crops. However, they have not been developed under specific variables and conditions [40–43]. However, protocols must be evaluated and adapted to each specific crop for practical application and for the rapid extraction of image features that represent pasture yield traits [44]. In this context, the purpose of this study is to develop precise and efficient methods to estimate oat yield using the germination index, agronomic variables, and spectral indices obtained through UAVs equipped with multispectral cameras, all through robust predictive models. These methods will enable the development of innovative solutions and, in the medium to long term, improve the efficiency of forage production while maximizing both the yield and quality of crops. However, it is important to highlight that the originality of this study lies in its focus on obtaining effective data through unmanned aerial vehicles under the specific conditions of the Peruvian Andes, a perspective that has not been Remote Sens. 2024, 16, x FOR PEER REVIEW 3 of 21 Remote Sens. 2024, 16, 3720 vehicles under the specific conditions of the Peruvian Andes, a perspective that3 hofa2s1 not been sufficiently explored in previous studies related to forage in these regions. Therefore, emsupffihcaiseinztilnyge txhpislo orreidgiinnalpitryev isio cursucsitauld tioe surnedlaetresdcotroe ftohrea gueniiqnuteh ecosentrreigbiuotnios.n Tohf ethreisfo srteu,dy. emphasizing this originality is crucial to underscore the unique contribution of this study. 2. Materials and Methods 2.21.. MExapteerriimalesnatnald SMitee thods 2.1.TEhxep efiriemlden etxapl Seirtiement was conducted at the Santa Ana Agricultural Experimental Cen- ter (heTrheaeftfieerl dreefexrpreerdim toe nats wSaans tcao Andnuac),t epdaratt otfh tehSea NntaatiAonnaal AIngsrtiictuultteu orafl AEgxrpaerriiamn eInntnaol va- tioCne n(ItNerIA(h)e (r7e5a°f1te3r′1r7e.f6e0r″reWd, t1o2°a0s′4S2a.nta◦ ′ ′′ 36″◦S A),n lao)c,ap′ t art of the Na ′e′d in the Man ttiaornoa Vl aInllsetyit uinte thoef cAegnrtararila hnigh- laInndnso voaf tion (INIA) (75 13 17.60 W, 12 0 42.36 S), located in the Mantaro Valley in thecentral hPiegrhula. nSdasntoaf APenrau .isS asnittuaaAtenda iast saitnu aatletidtuadt ea nraanltgitiundge frraonmgi n33g0f3ro tmo 33332035 tmo 3a3b2o5vme sea leavbeol.v Tehseea rleegvieoln. T’sh celirmegaitoen i’ss cclhimaraatcetiesrcizheadra bctye rai zreadinbyy saearasionny fsreoamso nNforovmemNboevr etmo bMerartoch, a trManasricthio,naatlr apnhsaitsioe nfarol pmh aAsperfirlo tmo AOpcrtoilbteorO, actnodb ear ,darnyd saedasryons efarsoomn fMroamy Mtoa Ay utogAusutg, uwsitt,h a towtaitl hanantoutaall panrencuiaplitparteiocinp iotaf t4io7n7 omfm47 7[4m5]m. A[v45e]r.aAgev etreamgepetermatpuererast urarnesgrea fnrgoemf r3o.m903 t.o90 20.2 °Cto, w20i.t2h◦ tCh,ew loitwhetshte tleomwpeestratetumrpeesr antudr efrseaqnudefnrte qfruoesntst forcocsutsrroicncgu rbreintwg ebentw Meeany Manady Aanudgust [4A6,u4g7u].s t [46,47]. TThhee eexxppeerriimeennttaall work was carrrriieed oouutti ninp plolotst,s1, 515m mlo lnognbgy b5y m5 mwi dweid(Fei g(Fuirgeu1r)e, 1), eaecahch susubbddivividideedd iintto fifive ssttrriipss.. FiFgiugruer e1.1 L. Loocacatitoionn ooff tthhee fifielld experriimeenntta anndde exxppereirmimenetnatladl edseigsnigonf osifx slioxc laolcoaalt ovaatr vieatrieiestiinesS ain tSaanta AAnan,a s, hshoowwiningg tthhee ggrroounnd ccoonnttrroollp pooinintsts( G(GCCPPs)s.). TThhee ooaat tlilnineess tteesstteed iin tthee eexxpeerriimeennttw weerereI NINIAIA2 020000a0n adnSdA SNATNATAN AAN, Abo, tbhoothri goir-igi- nantaitningg inin NNoorrtthh AAmeerriiccaa,, aanndd iinnttrroodduucceeddf oforrg grarianinp rpordoudcuticotnio. nA. fAtefrtearp ar opcreoscseosfsg oefn geetincetic imimpprorovveemmeennt taanndd pphheennoottyyppiicc sseelleeccttiioonni nint htheeM ManantatraoroV aVlalelyle, yso, msoemseh oswhoewd epdo tpenottieanltfioarl for fofroargagee pprroodduuccttiioonn.. TThheeiirr hhoommooggeneneietyityh ahsabse beneesntu sdtiueddifeodr fyoera ryse, astrasn, dstaarnddizainrdgiczhinarga ccthear-rac- teirsitsictiscssu scuhcahs asisz esi,zyeie, lydi,edldis,e dasiseeraesseis traenscisetaanndcea danapdt aabdilaiptytatobidliitfyfe rteon dt iafflteitruednets a. lTtihteudlaettse. rThe lagtteenre rgaetnederfiavteeds ufibvlien essucbalillneeds ScaanltlaedA nSaanshtao rAt mnae dium size (SA-MB), Santa Ana early tallsize (SA-PA), Santa Ana late medium size (SA-T),sShaonrtta mAendaieuamrl ysilozew (sSiAze-M(SAB)-,P SBa),natna dAna eaSralnyt ataAll nsaizme e(dSAiu-mPAh)i,g Shasntatatu Aren(aS Ala-tMe mA)e(dIiNuImA s2i0z0e0 )(.SA-T), Santa Ana early low size (SA- PB), and Santa Ana medium high stature (SA-MA) (INIA 2000). Remote Sens. 2024, 16, x FOR PEER REVIEW 4 of 21 Remote Sens. 2024, 16, 3720 4 of 21 2.2. Methodological Framework Figure 2 shows the comprehensive methodological framework used in this study, pro2v.2id. Minegth aondo olovgeicravliFerwam oefw tohrke sequential and integrative processes involved. The study is dividedF iignutore s2evsheorawl sktehye sctoamgpesre: htehnes ifivrestm ienthvoodlvoelos gtihcael cfroalmleecwtioonrk ouf saedgrionntohmisisct uddayt,a, fol- lowperdov bidyi nthgea nacoqvueirsviiteiwono of fth ienfsoeqrmueanttiioanl a tnhdroinutgeghr aUtiAvVe p flroigchestsse esqinvolved. The study isdivided into several key stages: the first involves the collection of agrouniopmpeicdd watait,hfo all omwueldtispec- tralb ycathmeearcaq.u Tishiteio tnhoirfdin fsotarmgea tieonnctohmropuagshseUsA tVhfle igehxttsraecqtuioipnp eodf wspitehcatrmalu iltnisdpieccetsr aalncadm p-hoto- graemram. Tehtreicth pirrdoscteasgseese,n caonmdp fiasnsaelslyth, ethexet rtaractiinoin ogf aspnedc tvraalliidndaticioesna onfd prheodtoicgtriavme meotrdicels are conpdroucetsesde.s , and finally, the training and validation of predictive models are conducted. FiguFirgeu 2re. 2D.eDscersicpritpiotino noof fththee methodollooggicicaal lf rfarmameweworokrekm epmlopyleodyiendt hinis trhesise arrecshe;aDrcShM; D(DSigMit a(lDigital SurSfaucrefa MceoMdoedl)e;l )D; DTMTM (D(Digigiittaall TerraiinnM Mooddele)l;)D; DHMHM(D (igDitiaglitHael iHghetigMhotd Mel)o.del). 2.3. Image Acquisition and Preprocessing 2.3. Image Acquisition and Preprocessing The experiment was conducted using a multispectral camera MicaSense Red Edge P (SeTahttele e, xWpAe,rUimSAen) tm wouanst ceodnodnuacDteJIdM uastirnicge a30 m0 RuTltKisUpAecVtr(aDlJ IcTaemchenrao lMogiycCaSoe.,nLsted .R, Sehde nE-dge P (Sezahttelne,, CWhAin,a U). STAhe) cmamouernatetadk eosn1 6a- bDitJmI Multaitsrpieccet r3a0l0d iRgiTtaKl iUmAagVe s(iDnJfiI vTeescihmniloalrosgpye cCtroa.l, Ltd., Shebnaznhdesn—, bCluhein(4a7)5. ±Th3e2 ncamm),egrrae etnak(5e6s0 1±6-2b7itn m)u, lrteidsp(6e6c8tr±al 1d4ingmita),l 1Ni4mIR5a6(g 8e4 s21 0±in8 48fi0vnem s),imilar speacntdraRl Eba(7n1d7s±—1b2lnume )(—47w5 i±th 3a2r ensmol)u,t giorneeonf 1 (.566m0e ±g a2p7i xnemls )(,1 4r5e6dx (6160888 ±p 1ix4e lns)m; i)n, aNdIdRit i(o8n4,2 ± 40 nma),G aNndSS RrEec (e7iv1e7r ±(D 1-2R TnKm2)—Mowbiitlhe Sat arteisoonluDtJiIo, nC hoifn 1a.)6w maseguasepdixaeslsa (real-time kinem paitxicels); in add(RitTioKn) ,b aas eGsNtaStiSo nr.eFciegiuvreer3 (sDh-oRwTsKth2e Meqoubipilme eSnttautisoend fDorJIc,a Cpthuirninag) mwualsti supseecdtr aalsi maa rgeeasl-time kinienmthaetiecv (aRluTaKtio) nbpalsoet .station. Figure 3 shows the equipment used for capturing multi- spectral images in the evaluation plot. Remote Sens. 2024, 16R,e3m7o2t0e Sens. 2024, 16, x FOR PEER REVIEW 5 of 21 5 of 21 Figure 3. (A) DJIFRigTuKreV 32. G(AN) SDSJ,I (RBT)KU AVV2 GMNaStrSi,c (eB3)0 U0,A(VC )MMatirciacsee 3n0s0e, R(Ced) MEdicgaesePnscea mRedra E, d(Dge) Pfl icgahmt era, (D) flight plan, (E) ground pcolannt,r o(El )p gorionutn(Gd CcoPn),tr(oFl) peovianltu (aGtiCoPn)p, (lFo)t ,eavnadlu(aGti)onC aplliobtr,a atnedR (eGfl)e CctalnibcreaPtea nRelfl(eCcRtaPn)c.e Panel (CRP). The flight planTwhea fls iegxhetc pultaend waarso uexnedcuntoedon arloucanldt nimooen, alotcaalh teimigeh,t aat ba ohveeigghrto aubnodveo gfround of 150 150 m. Images wme. rIemtakgens weveerrey t2a.k0esn wevitehry7 52%.0 fsr ownitthan 7d5%sid feroonvte arnladp .siTdhee osevedrilgaipta. lTihmeaseg edsigital images were stored in 1w6e-breit s.ttoifrfefdo rinm 1a6t-.bit .tiff format. The process begTahne wpritohcegseso bloegcatnin wgithe giemolaogceasticnagp thuer eidmbagyetsh ceaUptAuVreedq buyip tphed UwAiVth eaquipped with multispectral sean msour,lteinspsuecrtirnagl sthenesaocrc,u ernascuyroinfgt htheeg aecocgurraapchyi cofc othoer dgienoagteraspohfitch ceoiomrdagineast.eSsi oxf the images. ground control Spioxi ngtrsowunedre cuosnetdroals pfioxinedtsr wefere nucseesdo anst hfiexetedr raeifner, eanllcoews ionng tfhoer ctoerraeicnti,o anllofwing for cor- any deviation inretchtieopn oosfi tainony dofevthiaetiomna igne sthcea ptousriteidonb yoft htheeU imAVag. Oesn ccaeppturerecids ebyg etohleo UcaAtiVo.n Once precise was establishedg,eionldoicvaitdiouna wl iams aegsetasbwlisehreda,l ignndeivdidtouafol irmaagecso wheerreen atlmignoesadi cto, afocromn tain cuoohuersent mosaic, a and unified repcreosnetnintautoiouns oafntdh eunstiufidedy arerepar,ewsehnetaretiovne orlfa tphse wsteurdeyre amreoav,e wdhaenrde poevresrplaepctsi vweere removed differences weraenad jpuesrtespde. cPtrivecei sdeiffaelirgenmceesn wt iesrcer audcijualsteode.n Psruerceisteh atliaglnl mgeeongtr iasp chriuccfieaal ttuor esnsure that all are accurately rgeeporgersaepnhteicd fienatuhreems aorsea aicc.curately represented in the mosaic. Multispectral imMauglteispweectrreacl aimptaugresd wuesrien gcatphteuMredic auseingse thRee Md EicdasgenPsec Ramede rEadfgreo mP caanmera from an approximate heaigphptroxfi4m0amte, hweitghhat nofa 4v0e mra,g we iothf 1a4nfl aivgehrtasgceo onfd 1u4c ftleigdhetsv ecroynd7udcatyeds. eTvheeryU 7A dVays. The UAV flight time wasfalipgphtr otixmime awtaesly a0p4prmoxinimaantdely15 04s tmoicno avnedr 1th5e s eton tcioreverx tpheer iemnteinret aelxfipelrdimaerneata.l field area. During preprocDeussriing, prraedpioromcestrsicngca, rliabdriaotmioentrainc dcacloibrraetcitoionn anodf tchoerriencftoiormn aotfi tohne winefroermation were carried out. Thciasrrsiteedp owuat.s Tehsisse nsteiapl wtoasim epssreonvteiadl atota imquparloivtye bdyatad qjusatlintyg bvya raidatjuiosntisngin variations in lighting and atmligohsptihnegr iacncdo nadtmitioosnpshtehraict caondaiftfieocntst htehatc cuanra cayffeocftm theea saucrceumraecnyt socfa mpteuarseudrements cap- by the sensors. tRuareddio bmy etthreic secnalsiobrrsa.t Rioandieonmsuertreidc cthalaitbraetflioecnt aencsuerveadl utheasti rnefltheectiamnaceg evsalwuerse in the images consistent and wcoemrep caornasbilset.ent and comparable. PhotogrammetrPihcoptroogcreasmsimngetwriacs ppreorcfeosrsminegd wusaisn gpePrifxo4rDmePdro uMsinapg pPeirx4vD4. 8P.4r.os Moftawppareer v4.8.4. soft- (Prilly, Switzerwlaanrde ),(Pgreilolym, eStwriictzalelrylacnodr)r,e gcteeodmwetirtihcafilleyl dcoGrrCecPtesdi nwtoithth fieepldro GceCsPsisn igntfloo twhe processing to enhance thefltopwo tgor aenphaicncaec ctuhera tcoypogfrtahpehpico ainctcucrloaucyd oafn tdhet hpeoirnetfl celcotuadn caendb atnhde sreoflfetchtaence bands of orthomosaic, wtihthe aorfitnhaolmGorsoauicn,d wSiathm ap filinagl DGirsotaunncde S(GamSDpl)ionfg2 D.8isctman.cFer o(GmSDth)e opf o2i.n8t ccmlo.u Fdr,om the point a Digital SurfacceloMuodd, eal D(DigSiMtal) Swuarsfagceen Meroadteedl (wDiSthMt)h we assa mgeenreersaoteludt iwonitha sththee soamrtheo rmesoslauitcion as the or- and exported inth.otimffofosarmic aatn.d exported in .tiff format. Subsequently, aSuhbigsheq-rueesnotlluyt,i ao nhiogrhth-roemsoolusatiiocnw oarsthgoemnoersaatiec dw,agse goemneetrraitceadll,y gecoomrreectrtiecdally corrected to maintain a utnoi fmoramintsacianl eat hurnoiufogrhmo ustciatlse etxhtreonut,gehloimuti nitast inexgtednits,t oerltimioninsactianugs eddisbtoyrptioern-s caused by Remote Sens. 2024, 16, 3720 6 of 21 spective and topography. This allowed for precise measurements of distances, areas, and other features directly on the image. Finally, spectral index maps were generated, such as the NDVI (Normalized Difference Vegetation Index), which are valuable tools for analyzing vegetation health and vigor. These indices were applied in various prediction algorithms, enabling informed decisions linked to precision agriculture. 2.4. Field Data Acquisition The experimental work was carried out between December 2022 and July 2023. Twenty- four experimental plots measuring 16 m in length by 5 m in width were used, each subdi- vided into five rows. Six varieties of oats were selected based on their earliness and size, with four replications per variety. A total of 50 seeds were planted in each row, totaling 250 seeds per plot. Agronomic management followed conventional standards, including soil preparation with disc plowing and harrowing. Planting was carried out in rows, arranged in five rows per treatment, with 50 oat seeds planted in each row, spaced 30 cm apart. Since planting coincided with the rainy season, additional irrigation was not necessary. Manual weeding was performed, and 150 kg of urea fertilizer was applied. Evaluations were conducted every thirty days throughout the experimental period. Additionally, six georeferenced concrete control points were installed in the surrounding area of the experimental plot. In each plot, sampling points were marked every two meters along its length, specifi- cally selecting the three central rows for evaluation. Biometric measurements of oat plants in the central rows were taken, with a total of 21 samples per plot. Plant height measure- ments were conducted from March to July 2023, using a 2 m aluminum ruler for data recording. Plant survival was assessed on four occasions in December 2022 and January and March 2023, applying the survival percentage formula [48]. (Number of surviving plants) % Survival = × 100 (1) (Number of plants planted) A flower count was conducted in March 2023. During the harvest on 25 July 2023, dry biomass was weighed using an analytical balance, and the number of tillers per plant was counted. In the post-harvest phase, seed weight was evaluated both per plant and per plot, proceeding with manual separation of straw from seeds and using a sorting machine for this purpose. In the laboratory, two germination tests were conducted. The first test took place in April 2023 in an incubator at 25 ◦C, using four replicates of 100 seeds. Evaluations were conducted at 3, 7, and 10 days. The second test was conducted in July 2023 in a germination chamber at 25 ◦C under constant light and dark conditions, using trays with 72 holes filled with sterile peat. These evaluations were also conducted at 3, 7, and 10 days. Results were expressed as percentages of normal seedlings, abnormal seedlings, and hard, fresh, and dead seeds using the methodology described in [48]. (Total seeds planted) % Germination = × 100 (2) (Total seeds germinated) A total of six data collection sessions were conducted during the growing season, spanning from the early stem elongation stage to the late senescence stage. These field measurements were taken in December (14, 21) of 2022 and January 2023 (6,18), as well as on 22 February and 8 March 2023, on the same days as the studies and UAV flights conducted in December (14, 21, 28) of 2022, as well as January (6, 11, 18, 25), February (8, 10, 16, 22), and March (2, 8, 15) of 2023, totaling 14 flights to provide real-time ground data. Remote Sens. 2024, 16, 3720 7 of 21 2.5. Extraction and Processing of Multispectral Images For the processing and extraction of multispectral data, the photogrammetric process was initiated using Pix4D Pro Mapper software (Prilly, Switzerland). This software is essential for generating detailed orthomosaics from multiple images captured by the multispectral camera mounted on the UAV. This process included not only the generation of digital elevation models but also the fusion of spectral data to enhance the spatial and thematic resolution of the resulting images. The R studio software (R Core Team) was used for the analysis and manipulation of the obtained geospatial data. Additionally, the Terra package for Hijmans [49] was employed for image processing and the extraction of spectral indices, and Quantum Geographical Information System software (QGIS 2.18.14, QGIS Development Team, Raleigh, NC, USA) was used for the vectorization of the images, enabling a more detailed analysis of the study plots. For the extraction of indices, the derived spectral bands were used based on biblio- graphic equations from the Terra package [49]. With the indices calculated, a 30 cm buffer was created around the central point of each plant, and within this buffer, the zonal statistics of the maximum values of all pixels contained within the buffer were extracted. To develop the benchmark models, we initially generated a set of 21 spectral indices. These indices included vegetation, soil, and water indices, and they are closely related to crop biomass. For each sampled point, a circular buffer with a radius of 0.25 m was used. The vegetation indices were calculated through different combinations of reflectance and compiled as predictors along with the pure spectral bands (Table 1). Table 1. Spectral indices derived from UAV-acquired multispectral images. Indices Equation Description Differenced Vegetation Index (DVI) Nir − Red [50] Normalized Difference Vegetation Index (NDVI) Nir−Red [51] Nir+Red Green Normalized Difference Vegetation (GNDVI) Nir−GreenNir+Green [52] Normalized Difference Red Edge (NDRE) Nir−ReNir Re [52]+ Enhanced Normalized Difference Vegetation Index (ENDVI) 2×Nir−Red−Blue× [53]2 Nir+Red+Blue Renormalized Difference Vegetation Index (RDVI) √Nir−Red [54] Nir+Red Enhanced Vegetation Index (EVI) G × (Nir−Red) [55] (Nir+C1×Red−C2×Blue+L) Visible Difference Vegetation Index (VDVI) Nir−Red [56] Nir+Red+Green Wide Dynamic Range Vegetation Index (WDRDVI) √(Nir−Red) [57] Nir+Red Transformed Vegetation Index (TVI) 0.5 × [120 × (Nir − Green)] [58] Soil-Adjusted Vegetation index (SAVI) Nir−Red [59] Nir+Red+L × (1 + L) Optimized Soil-Adjusted Vegetation Index (OSAVI) (Nir(−Red) ) [59]Nir+Red+0.16 Content Validity Index (CVI) Nir × Red [60] √ Green Modified Soil Adjusted Vegetation Index (MSAVI) 2×Nir+1 (2 x Nir+1)2−8×(Nir−Red) [61] 2 Modified Chlorophyll Absorption in Reflectance Index (MCARI) (Red−Green)−2×(Red−Blue) [62] Red Green Transformed Chlorophyll Absorption in the Reflectance Index (TCARI) 3 × ((Red − Green)− 0.2 × (Red − Blue)× ( RedGreen )) [63] Normalized Pigment Chlorophyll Reflectance (NPCI) Red−Blue [60,64] Red+Blue Green Coverage Index (GCI) NirGreen − 1 [56] Red-Edge Chlorophyll Index (RECI) NirRe − 1 [65] Structure Insensitive Pigment Index (SIPI) Nir−Blue− [66]Nir Red Anthocyanin Reflectance Index (ARI) 1 − 1Green Red [67] Multispectral imagery central wavelengths: Blue, Green, Red, Red edge (Re), and Nir: 474, 560, 668, 717, and 840 nm. The selection of these indices was not random. Each index was chosen based on its theoretical and empirical capacity to reflect key aspects of crop condition and health, such as biomass, soil coverage, and plant water content, which are critical for accurate yield estimation. We initially calculated a broad range of indices to ensure that we covered all potential variables that could influence oat yield. Subsequently, we applied statistical techniques to select the two most significant indices, which provided valuable information for the predictive models. Remote Sens. 2024, 16, 3720 8 of 21 2.6. Data Analysis and Selection of Predictor Variables For the analysis of agronomic data, homoscedasticity tests and normality tests [68] were conducted to determine the distribution and symmetry of the data. Additionally, box- plots were constructed to identify outliers, enabling efficient data cleaning and refinement. Spectral indices were extracted for each flight date, totaling 28 flights for oat crop monitoring. Twenty-one spectral indices were evaluated in each flight. Subsequently, a correlation matrix was generated between agronomic variables and spectral indices derived from multispectral images, considering flight dates and days elapsed since planting, ranging from 20 to 250 days. Out of the 21 indices evaluated, only 2 exhibited suitable behavior or correlation for further processing: NDVI and NDRE. Furthermore, a significance matrix was developed to identify indices showing higher Pearson’s correlation with agronomic variables. It was observed that data obtained beyond 100 days post-planting exhibited stronger correlations with agronomic variables measured in the field. The process included correlating variables derived from spectral indices with agronomic parameters measured using traditional techniques such as height, germination percentage, flower count, grain dry matter, stem count, survival percentage, and dry biomass weight. This correlation was crucial for identifying significant relationships and understanding complex patterns affecting biomass estimation. Using this information, Principal Component Analysis (PCA) was performed consid- ering only days post-planting exceeding 100. Subsequently, a detailed correlation analysis was conducted to identify spectral indices most strongly related to agronomic variables. This approach helped identify critical periods and the most relevant spectral indices. These tools enabled clustering of areas with similar characteristics and reduced data dimensional- ity, facilitating identification of the most relevant variables. Identification of key variables for the study was accomplished through a Pearson correlation matrix. This matrix identified variables significantly correlated with the variable of interest, namely dry biomass weight and the most representative flight days. 2.6.1. Modeling and Estimation Algorithms We use four predictive regression algorithms. First, linear regression models the rela- tionship between independent variables and a dependent variable through a straight line, which is useful for predicting numerical value [69–71]. Second, Random Forest combines multiple decision trees to improve prediction accuracy and reduce overfitting [72–74]. Third, Neural Networks (NNs) are brain-inspired models composed of layers of interconnected nodes that process information and learn complex patterns through feed- back [75]. These models are suitable for deep learning and complex pattern recognition problems [76,77]. Finally, Support Vector Machines (SVMs) are supervised learning algorithms that seek the optimal hyperplane to separate data into different classes in a high-dimensional space [78]. They are effective on both small and large datasets and can be applied to linear and nonlinear classification problems. Although they may face computational challenges with very large datasets, optimization techniques and advanced computational resources can enhance their efficiency [79,80]. 2.6.2. Model Tuning and Evaluation The response variable was selected as a vector, while the predictor variables were grouped into a matrix, with the data split into 70% for training and 30% for testing. The split was performed randomly and not based on different time periods. To optimize the performance of the machine learning models, it is essential to fine-tune their hyperparameters. We used the training samples to calibrate the models using the Grid Search method [81,82] which exhaustively evaluates all possible combinations of hyperparameters. Ten-fold cross-validation and the coefficient of determination (R2) were employed as evaluation metrics to ensure the robustness of the models [83,84]. Remote Sens. 2024, 16, 3720 9 of 21 For Random Forest, the hyperparameters ‘ntree’ (testing values between 20 and 180) and ‘mtry’ (testing values between 2 and 14) were adjusted, with other parameters kept at their default settings. For the Support Vector Machine (SVM), ‘C’ (testing values between 0.01 and 10) and ‘Gamma’ (testing values between 0.1 and 1) were adjusted, while the remaining parameters were also set to their default values. The architecture of the neural network used has two hidden layers with 6 and 5 neu- rons, respectively, with an input layer with the number of nodes equal to the number of predictive variables and the output layer with one neuron and a linear continuous response. The modeling was carried out in R, using the ‘randomForest’, ‘e1071’ and neuralnet libraries [85]. 3. Results 3.1. Descriptive Statistics of Agronomic Variables Multiple metrics reflecting oat crop development were evaluated. Various descriptive statistics were analyzed (Table 2) for the parameters of each evaluated variable. Table 2. Descriptive agronomic statistics of oat variables. Agronomic Variable Prom Mean Min Max σ Height (meters) (h) 1.50 1.50 1.13 1.88 0.15 Tillers (t) 29.97 29.00 5.00 61.00 11.57 Dry matter (kg) (dm) 0.29 0.28 0.03 0.63 0.12 Grain weight (kg) (gw) 0.04 0.04 0.00 0.10 0.02 Flowers (f) 14.76 16.00 3.00 26.00 5.38 Germination percentage (gp) 0.72 0.79 0.30 0.88 0.18 Germination in emergency chamber (gem) 0.62 0.71 0.15 0.78 0.18 Total plants (tp) 31.87 33.00 15.00 40.00 5.47 Survival rate (sr) 0.64 0.66 0.30 0.80 0.11 Germination rate (gr) 10.27 11.29 4.29 12.57 2.55 The oat at harvest reached an average height of 1.50 m, with a standard deviation of 0.15 m, and a range varying between 1.13 and 1.88 m. The number of stems per plant had a mean of 29.97, with a standard deviation of 11.57, ranging from 5 to 61 stems. Regarding dry matter, plants averaged 0.29 kg, with a standard deviation of 0.12 kg and a range from 0.03 to 0.63 kg. Grain yield per plant averaged 0.04 kg, with a standard deviation of 0.02 kg and a range from 0.00 to 0.10 kg. Additionally, an average of 14.76 flowers per plant was observed, with a standard deviation of 5.38, and a range fluctuating between 3 and 26 flowers. Germination data showed an average percentage of 72%, with a standard deviation of 18% and a range from 30% to 88%. 3.2. Spectral Variable Analysis 3.2.1. Significance Correlation Matrix The correlation matrix (Figure 4) displays Pearson correlation coefficients between agronomic variables and filtered spectral indices, calculated using the Corrplot library for Wei T [86], reflecting their linear relationships. Matrix values range from −1 to 1, where 1 indicates a perfect positive correlation, −1 is a perfect negative correlation, and 0 is no correlation. The main diagonal of the matrix always holds 1, as each variable correlates perfectly with itself. Remote Sens. 2024, 16, x FOR PEER REVIEW 10 of 21 The oat at harvest reached an average height of 1.50 m, with a standard deviation of 0.15 m, and a range varying between 1.13 and 1.88 m. The number of stems per plant had a mean of 29.97, with a standard deviation of 11.57, ranging from 5 to 61 stems. Regarding dry matter, plants averaged 0.29 kg, with a standard deviation of 0.12 kg and a range from 0.03 to 0.63 kg. Grain yield per plant averaged 0.04 kg, with a standard deviation of 0.02 kg and a range from 0.00 to 0.10 kg. Additionally, an average of 14.76 flowers per plant was observed, with a standard deviation of 5.38, and a range fluctuating between 3 and 26 flowers. Germination data showed an average percentage of 72%, with a standard deviation of 18% and a range from 30% to 88%. 3.2. Spectral Variable Analysis 3.2.1. Significance Correlation Matrix The correlation matrix (Figure 4) displays Pearson correlation coefficients between agronomic variables and filtered spectral indices, calculated using the Corrplot library for Wei T [86], reflecting their linear relationships. Matrix values range from −1 to 1, where 1 indicates a perfect positive correlation, −1 is a perfect negative correlation, and 0 is no Remote Sens. 2024, 16, 3720 correlation. The main diagonal of the matrix always holds 1, as each va1r0iaobf l2e1 correlates perfectly with itself. Figure 4. CorreFliagtuioren 4c. oCeoffirrceileantitosnb ceotewffieceinenatgs rboentwomeeinc avgarroinabomlesic avnadriasbpleecst arnaldv sapreicatbralel svaorviaebrletism oev.er time. r— r—Pearson corrPeelaatrisoonn ccooerrffielcaiteinont; csoigenffiificiceannt;t saitgnthifiec5a%nt part othbea b5i%lit pyrloebvaebl;ilXity= lenvoetl;s iXg n= infiocta snitg.nificant. This analysis was crucial for identifying patterns, detecting multicollinearity, and This analsyesliesctwinags tchreu mcioaslt froerleivdaenntt vifayrinabglepsa ctotenrcnesrn, idnegt edcrtyin bgiomuaslsti wcoelilginhet a(briwty),, wanitdhout seeds. selecting the mIno stthree cleovrraenlattvioanri anbalelysscios nwcietrhn dinryg mdrayttberi o(mdmas) sasw tehieg hvat r(ibawbl)e, owf iitnhtoeuretsste, evdarsy. ing levels In the correlatiofn aasnsoaclyiastiisown iwthitdh rsyevmearattl evra(rdiambl)eass wthereev faoruianbdl.e of interest, varying levels of association with several variables were found. The correlation coefficient between dry matter (dm) and NDRE_160 was −0.1, indicat- ing a weak negative correlation. NDVI, measured on days 111, 131, 141, and 146, showed correlation values ranging from 0.15 to 0.21 with dry matter, suggesting a weak to moderate positive correlation. Similarly, agronomic variables exhibited a positive correlation of 0.2 with dry matter (dm). In contrast, the number of stems displayed a strong positive correlation of 0.76 with dry matter, suggesting that a higher number of stems is strongly associated with an increase in dry matter. Additionally, the grain weight showed a positive correlation of 0.2 with dry matter. The flight days that showed the highest correlation were 111, 118, 125, 131, 141, 146, 153, 160, and 167. 3.2.2. Principal Component Analysis Principal Component Analysis (PCA) enabled a reduction in data dimensionality by eliminating redundancies and focusing the analysis on the most influential variables. This reduction not only simplified the complexity of the dataset but also decreased both squared and absolute errors in the predictive models, thereby enhancing their accuracy and efficiency. Consequently, PCA has established itself as a key tool for optimizing the performance of machine learning algorithms in biomass estimation and crop yield prediction [87–90]. Figure 5 shows PCA, where Principal Component 1 (PC1) explains 30.21% of the variance and Principal Component 2 (PC2) explains 15.17%. Each arrow represents an Remote Sens. 2024, 16, x FOR PEER REVIEW 11 of 21 The correlation coefficient between dry matter (dm) and NDRE_160 was −0.1, indicating a weak negative correlation. NDVI, measured on days 111, 131, 141, and 146, showed correlation values ranging from 0.15 to 0.21 with dry matter, suggesting a weak to moderate positive correlation. Similarly, agronomic variables exhibited a positive correlation of 0.2 with dry matter (dm). In contrast, the number of stems displayed a strong positive correlation of 0.76 with dry matter, suggesting that a higher number of stems is strongly associated with an increase in dry matter. Additionally, the grain weight showed a positive correlation of 0.2 with dry matter. The flight days that showed the highest correlation were 111, 118, 125, 131, 141, 146, 153, 160, and 167. 3.2.2. Principal Component Analysis Principal Component Analysis (PCA) enabled a reduction in data dimensionality by eliminating redundancies and focusing the analysis on the most influential variables. This reduction not only simplified the complexity of the dataset but also decreased both squared and absolute errors in the predictive models, thereby enhancing their accuracy and efficiency. Consequently, PCA has established itself as a key tool for optimizing the performance of machine learning algorithms in biomass estimation and crop yield Remote Sens. 2024, 16, 3720 prediction [87–90]. 11 of 21 Figure 5 shows PCA, where Principal Component 1 (PC1) explains 30.21% of the variance and Principal Component 2 (PC2) explains 15.17%. Each arrow represents an original variabolerigpirnoajel cvtaerdiaobnlet optrhoejepctreidn coipntaol cthoem pproinnceinptasl pcaocmep; tohneendti rsepcaticoen; tahned dlierencgttihono af nd length the arrows indoifc athtee haorrwowths einsedivcaartiea hbolews cthoenster ivbaurtieabtloest hceonctormibpuoten teon tths.e components. Figure 5. PrincipFiagluCroe m5.p Porninecnipt aAl nCaolmyspisonoefnatg Aronnaolymsisc oafn adgsropneoctmraicl vanardi asbpleecst.ral variables. Variables NDVRaEr_ia1b6l7esa nNdDNRED_R16E7_ 1a6n0d aNreDnReEg_a1t6i0v ealrye cnoergraetliavteldy wcoirtrhelPatCe1d, wihthe rPeCas1, whereas NDVI variableNsD(V11I 1v,a1r1ia8b, l1es2 5(,1113,1 1, 1184, 11,2154, 61,315, 31,41,6 01,461,6 175)3c,o 1n6t0r,i b1u67te) csoignntriifibcuatne tslyigannifidcantly and positively to PpCo1s.itOivtehleyr tov aPrCia1b. lOesthseurc vhaarisabgleersm su, gchp ,ags rg,esrrm, t,p g,pt, grw, ,srh, rt,pa,n t,d gdwm, hcro, anntrdib dumte contribute to PC2 to varytoin PgCd2e gtor eveasr.yTinhge dceoglorereos.f Tthe acorrlorw osf rtehflee acrtrsotwhse iresfliegcntsifi tchaenirc esiglenvifielc,awncieth level, with darker colors idnadrikceart icnoglogrrse iantdeircasitginngi figcraenatceer. significance. In particular, the dry matter variable “dm” projects primarily in the positive direction of PC2 and slightly in the negative direction of PC1, suggesting a positive correlation with Principal Component 2 and a slight negative correlation with Principal Component 1. The darker intensity of the “dm” arrow indicates moderate to high significance. PCA reveals that NDVI variables strongly influence PC1, while other variables like germ and gp contribute more significantly to PC2. The variable “dm” stands out for its influence on PC2, correlating positively with other variables that also contribute to this component. 3.3. Model Performance Figure 6 shows two Taylor diagrams illustrating the performance of the models used to estimate dry matter: linear regression, Neural Network, Random Forest, and Support Vector Machine. Each point on the graph represents the accuracy of these models in terms of standard deviation, correlation coefficient, and root mean square error (RMSE) for both the training and test datasets. In the test set, the Support Vector Machine shows strong performance with a correlation coefficient (r) of 0.93, followed by the Neural Network model with an r of 0.91, indicating high precision in predicting dry matter (dm). Remote Sens. 2024, 16, x FOR PEER REVIEW 12 of 21 In particular, the dry matter variable “dm” projects primarily in the positive direction of PC2 and slightly in the negative direction of PC1, suggesting a positive correlation with Principal Component 2 and a slight negative correlation with Principal Component 1. The darker intensity of the “dm” arrow indicates moderate to high significance. PCA reveals that NDVI variables strongly influence PC1, while other variables like germ and gp contribute more significantly to PC2. The variable “dm” stands out for its influence on PC2, correlating positively with other variables that also contribute to this component. 3.3. Model Performance Figure 6 shows two Taylor diagrams illustrating the performance of the models used to estimate dry matter: linear regression, Neural Network, Random Forest, and Support Vector Machine. Each point on the graph represents the accuracy of these models in terms of standard deviation, correlation coefficient, and root mean square error (RMSE) for both the training and test datasets. In the test set, the Support Vector Machine shows strong Remote Sens. 2024, 16, 3720 performance with a correlation coefficient (r) of 0.93, followed by the Neural N1e2towf 2o1rk model with an r of 0.91, indicating high precision in predicting dry matter (dm). FFigiguurree 66. .TThhee TTaayylloorr ddiiaaggrraam compares the perforrmaanncceeo offl liinneeaarrr reeggreressisoionn( L(LMM),)N, NeueruarlaNl Netewtworokrk (N(NNN)), ,Random Foorreesstt( R(RFF),)a, nadndS uSpuppoprtoVrte cVtoecrtMora cMhianceh(inSVe M(S)VmMo)d emlsoidneplsr eidni cptirnegdidcrtyinmg adttreyr m(damtt)er (dbmas)e dbaosnesdt aonnd asrtdanddeavriadt idonev, ciaotriroenla, ticoonrrceolaeftfiiocnie ncot,eaffindcieRnMt,S aEnfdo rRbMothSEtr afoinr inbgotahn dtrtaeisntindgat aasnedts .test datasets. The Random Forest and Support Vector Machine models show outstanding fit in the tTrahien Rinagnsdeotmw iFthorresvta aluneds Soufp0p.7o0rta nVdec0t.o7r2 ,Mreascpheinctei vmeloyd, edlesm shoonwst roautitnsgtahnidgihngle fiartn inin tghe trcaipnaicnigty s. eHt owwitehv err ,vtahleuiersp oefr f0o.r7m0 aanncde 0in.7t2h,e rteessptescettiviselmy,o dreemodnestsrta. tOinvge rhailgl,ht hleaSrunpinpgo rcta- pVaeccittyor. HMoawcheivneer,m thoedierl peexrhfiobrimtsatnhceeb iens tthoev etreasltl speet rifso rmmoarnec me,ocdoemsbt.i nOivnegreaxllc, etlhleen Stufiptpinort Vtehcetotrra Minainchginseet mwoidthel seoxlhidibpitesr tfhoerm beasntc oevienratlhl epeterfsot rsmeta.ncOen, ctohme boitnhienrg heaxncedl,letnhte filtin inea trhe trraeginrienssgi osent mwoitdhe lsohlaisd tpheerlfoowrmesatnccoer rienl athtieo ntecsot esfefitc. iOennt tvhael uoethseinr hbaonthd,t htheet elisnteaanrd retrgariensinsigon mseotds,ehl ihgahsl itghhet ilnogwietsstl coowrererleaftfieocnti vcoeenffiescsiceonmt vpaalrueeds tion tbhoetoht htheer teevsatl uanatde dtraminoidnegls s.ets, high- lightinTgh eitSs ulopwpoerrt eVffeeccttoirveMnaecshs icnoemshpoawresdt thoe tbhees toothveerr aelvlapleuraftoerdm manocdeealss. it demonstrates low TRhMeS SEuipnpboortth Vdeacttaosre tMs aanchdinhieg shhrovwaslu tehse. Abedsdt iotivoenraalllly ,ptehrefoRramndanomce Faosr iets dt memodonelsatrlasotes lopwer fRoMrmSsEw ine blloitnht dhaettaessettsse atn, dal thhigouhg rh vanlouteass. Astdrodnitgiolynaalslyth, tehSeV RMan. dTohme N Feourersatl mNoetdwelo arklso pmerofdoerml ssh owwesll giono tdhefi tteinstt rsaeitn, ianlgthbouutglho wneort paesr sfotrromnagnlyce aisn ttheest SinVgM, s.u Tghgees Ntineguroavle Nrfietttwinogr.k mTohdeeliln sehaorwresg groession model performs the poorest, with the highest RMSE values and thelowest r values.od fit in training but lower performance in testing, suggesting overfitting. The linear regression model performs the poorest, with the highest RMSE values and the lo3w.4.ePstr erd vicatloureEss. ti mation In the results table of the linear regression (Table 3), coefficient estimates for var- ious predictors used in the model are presented. The intercept has a value of −0.439 (p = 0.020104), which is significant and suggests that, in the absence of other factors, forage biomass would have a negative value. The coefficient for NDVI_111 is 0.541 (p = 0.034539), indicating a positive and significant relationship with biomass. On the other hand, NDVI_125 has a coefficient of −1.148 (p = 0.012474), showing a significant negative relationship. NDVI_131 has a coefficient of 1.256 (p = 0.000604), indicating a positive and highly significant relationship. Finally, plant height (h) has a coefficient of 0.184 (p = 5.805), also significant, implying that an increase in height is associated with an increase in forage biomass. Remote Sens. 2024, 16, 3720 13 of 21 Table 3. Characteristics and estimation of predictors. Linear Regression Residual Coefficient Min −0.208 Intercept NDVI 111 NDVI 125 NDVI 131 h 1Q −0.091 Estimate −0.439 0.541 −1.148 1.256 0.184 Median −0.016 Std 0.188 0.256 0.457 0.363 0.045 3Q 0.075 Error −2.336 2.123 −2.513 3.464 4.075 Max 0.314 t-values 0.02 0.034 0.012 0.004 5.8 Parameters RF SVM NN Hyperparameters Node size 100 Cost 1 * HL 2 Neurons 5 Trees 500 γ 0.1 Neurons 3 N◦ var 4 ε 0.01 Residual Mean of squared 0.011 N◦ of vectors 294 % var 52.23 SVM Radial * 5% confidence Interval; RF (Random Forest), SVM (Support Vector Machine), NN (Neural Network); HL = hidden layer; 70% training set and 30% test set. The hyperparameters and metrics of predictive models provide crucial information about their configuration and performance. In the Random Forest model, a node size of 100 indicates that each terminal node must have at least 100 observations, with 500 trees improving accuracy but increasing server processing time, and four variables considered for splitting each node. For the Support Vector Machine model, the cost parameter is set to one, balancing between a wide margin and correct classification of training points; gamma of 0.1 indicates that only nearby points influence the decision boundary; epsilon of 0.01 reflects high precision, and 294 support vectors are used. In the Neural Network model, two hidden layers are configured with five and three neurons, respectively, affecting its ability to capture complex patterns. Regarding residual performance metrics, the mean squared residual is 0.011 for the Random Forest model, suggesting good accuracy, and the percentage of explained variance is 52.23%, indicating that this percentage of variability in the data is explained by the predictors used in the model, but also showing that 47.77% of the variability remains unexplained. The performance of different models shows variability in predictive capability. For linear regression, the coefficient of determination (R2) is 0.12 in the training set and 0.04 in the test set, indicating low explanatory power. Table 4 shows the performance of four models used (linear regression, Random Forest, Support Vector Machine, and Neural Networks) for estimating dry matter in the training and test datasets, evaluated using the metrics R2, RMSE, and MAE. Random Forest and Support Vector Machine stand out with superior results: in training, RF achieves an R2 of 0.68 with an RMSE of 0.080 and an MAE of 0.063, while SVM achieves an R2 of 0.87 with an RMSE of 0.047 and an MAE of 0.030. In testing, RF obtains an R2 of 0.68 with an RMSE of 0.080 and an MAE of 0.063, and SVM reaches an R2 of 0.83 with an RMSE of 0.051 and an MAE of 0.040, demonstrating robust generalization performance. Table 4. Performance of predictive models. Linear Regression RF SVM NN R2 0.12 R2 0.68 R2 0.87 R2 0.83 Training RMSE 0.117 RMSE 0.080 RMSE 0.047 RMSE 0.051 MAE 0.096 MAE 0.063 MAE 0.030 MAE 0.040 R2 0.04 R2 0.52 R2 0.50 R2 0.35 Test RMSE 0.119 RMSE 0.087 RMSE 0.085 RMSE 0.110 MAE 0.097 MAE 0.069 MAE 0.066 MAE 0.086 5% confidence Interval; RF (Random Forest), SVM (Support Vector Machine), NN (Neural Network). Remote Sens. 2024, 16, x FOR PEER REVIEW 14 of 21 Table 4. Performance of predictive models. Linear Regression RF SVM NN R2 0.12 R2 0.68 R2 0.87 R2 0.83 Training RMSE 0.117 RMSE 0.080 RMSE 0.047 RMSE 0.051 MAE 0.096 MAE 0.063 MAE 0.030 MAE 0.040 R2 0.04 R2 0.52 R2 0.50 R2 0.35 Test RMSE 0.119 RMSE 0.087 RMSE 0.085 RMSE 0.110 MAE 0.097 MAE 0.069 MAE 0.066 MAE 0.086 Remote Sens. 2024, 16, 3720 5% confidence Interval; RF (Random Forest), SVM (Support Vector Machine), NN (Neural Netw14oorkf )2.1 On the other hand, Neural Networks show better performance in training with an R2 of 0.8O3n, bthuet ostihgenrifihcaanndtl,yN deeucrraelaNsee tiwn otreksstinshgo wwibthe ttaenr pRe2r foofr m0.a3n5,c eininditcraatiinnign gpowtietnhtiaanl 2 2 oRveorfiftt0i.n83g, dbuurtinsigg ntriafiicnainngtl.y Ind eccornetarsaesti,n litneesatirn rgegwrietshsiaonn Rshoowf s0 .t3h5e, lionwdiecsatt ipnegrfporomteannticael wovitehr fiatnt iRn2g2 o df u0r.1in2 gantrda i0n.0in4g, a. nIn RcMonStEr aosft ,0.l1in1e7a arnrde g0r.1e1ss9i,o anndsh aonw Ms AthEe olof w0.e0s9t6p aenrdfo 0r.m09a7n icne twraitinhianng Randof t0e.s1t2inagn, dre0s.p04e,catinveRlyM. STEheo fS0u.1p1p7oartn dVe0c.1to1r9 ,ManadchainneM dAemE oonf s0t.r0a9t6esa ncdon0s.0is9t7enint ptrearifnoirnmgaanncde itnes btiontgh, trreasipneincgti vaenldy. tTeshteinSgu, pwpiothr taVne Rct2o2 o rfM 0.a8c7h ainnde d0e.5m0,o rnesstpreactetisvceolyn,s aislotenngt wpeitrhfo lromwa RncMeSinE banodth MtrAaiEn ivnagluaensd. testing, with an R of 0.87 and 0.50, respectively, along withTlohwe SRuMppSEorat nVdecMtoAr EMvaaclhuiense. proves to be the most effective model for this dataset, whileT lhineeSaur prepgorretsVsieocnto erxMhibacithsi mneoprer olivmesitetod bpeertfhoermmaonsctee finfe ccotimvepmaroisdoenl. for this dataset, while linear regression exhibits more limited performance in comparison. 33..55.. PPrreeddiiccttiivvee MMooddeell ffoorr BBiioommaassss EEssttiimmaattiioonn FFiigguurree 77 ddeeppicitcstst htehees teismtiamtiaotnioonf dorfy dmrya ttmeraittnefro rinag feooraatgseu osiantgs tuhseinevga tluhaet eedvamluoadteelds: mLMod(eLlsin: eLaMr M (oLdinele)a, rN NMo(Ndeelu),r aNl NNe t(wNoerukr)a, lR NF (eRtwanodrko)m, RFoFr e(sRt)a,nadnodmS VFMor(eSsut)p, paonrdt VSeVctMor (MSuapchpionret )V. eEcatocrh Mmaacphidnies)p. lEaaycsht hmeaeps tdiimspaltaioyns tihne0 e.2st5imma2tigorni dinc 0e.l2ls5, mco2l ogrreidd caeclclso, rcdoilnogretdo athcceoirrdwinegig htot inthkeiirlo wgreaimghst, ainll okwilionggrafomrso, baslelorvwaitniogn foofr tohbesvearvriaattiioonn oinf pthreed vicatrioiantisofnr oimn pearecdhimctioodnesl .from each model. Fiigurre 7.. Reeprressenttattiion off drry matttterr essttiimatted tthrrough prrediiccttiion modellss fforr oatt ccullttiivattiion.. LM predominantly predicts medium weights (0.10 kg to 0.30 kg), with few cells at extreme values. NN shows greater variability, with cells ranging from 0.05 kg to 0.55 kg, highlighting areas with higher biomass. RF exhibits a trend similar to LM, with average val- ues around 0.29 kg. In contrast, SVM shows a more balanced but conservative distribution, predominantly with lower weights (0.05 kg to 0.25 kg). Areas with gray outlines in all maps indicate regions with an estimated dry matter of zero due to the absence of vegetation. These maps allow for comparing the predicted biomass distribution by each model, facilitating understanding of their differences. In summary, SVM is the model that most closely approximates the average dry matter evaluated in the field, with an estimation Remote Sens. 2024, 16, 3720 15 of 21 ranging between 0.05 kg and 0.25 kg, demonstrating its effectiveness in estimating dry matter in forage oats. 4. Discussion The use of UAVs for biomass estimation has proven to be an efficient and accurate method, as indicated by a coefficient of determination (R2) of 0.52. Although this value is lower than those reported in previous studies, such as that by Lussem [91], who recorded R2 values between 0.56 and 0.73, it is important to note that the RMSE values in his study were considerably higher, ranging from 0.274 to 0.416. On the other hand, studies like Coelho’s [92], report significantly higher R2 values (0.70–0.89), which better capture data variability, albeit with higher absolute errors (RMSE from 370 to 1825 kg ha−1). In contrast, the present study reports an RMSE of 0.080 and MAE of 0.063, suggesting that while a model with a higher R2 may better explain variance, the higher absolute and squared errors could reduce its reliability due to sensitivity to deviations or outliers [93–95]. Therefore, balancing explanatory power and accuracy is crucial in the selection of estimation models. Despite these differences, the UAV approach continues to provide suitable and often superior estimates in many contexts [96,97]. Estimating biomass in crops through remote sensing involves several critical and interconnected objectives aimed at improving model accuracy and applicability. One of the primary goals is to identify the variables that correlate with reference biomass obtained in the field, which requires careful selection and evaluation of different sensors and remote sensing techniques [98]. Equally important is the development of accurate and scalable models that incorporate both parametric and non-parametric algorithms, and the integration of data from multiple sensors to enhance estimation precision [91]. However, it is crucial to recognize that the accuracy of these estimates can be affected by various sources of uncertainty. Environmental factors, such as light conditions, wind, and humidity, can significantly influence UAV image capture, highlighting the need for detailed analyses to identify and mitigate these potential errors [99]. Additionally, the spatial and temporal scale of the data plays a vital role, as different resolutions can significantly impact the accuracy of biomass models, necessitating constant methodological adjustments [98]. While the progress made is promising, applying these models on a larger scale presents additional challenges [100]. The transferability of these models across diverse geographic and temporal conditions must be rigorously evaluated to ensure their robustness and accuracy in large-scale applications [91,97,101]. In this context, it is essential to consider how environmental variability might affect the applicability of the results in different regions and under varying climatic conditions. Finally, to advance toward more efficient and precise methodologies in global agricul- tural management, it is crucial not only to enhance the accuracy of existing models but also to address the identified limitations, such as environmental factors and model scalability, in future research. Emphasizing the importance of a critical and thorough evaluation of the results ensures that the proposed solutions are viable and effective in a broader agricultural context. The use of machine learning models such as regression, Random Forest, Support Vector Machines, and Neural Networks for biomass estimation using UAVs has been extensively researched. Regression is useful for handling highly correlated predictors and has been effective in predicting biomass in various vegetation types, notably reduc- ing dimensionality and improving estimation accuracy [102]. RF is robust and accurate, successfully used in estimating biomass in different forests and crops, including forage oats, due to its ability to handle datasets with many features without overfitting [5,103]. The Support Vector Machine, capable of handling nonlinear and high-dimensional data, has also proven effective in biomass prediction using complex spectral data, although proper parameter selection can be challenging [104]. Neural Networks, on the other hand, can model complex nonlinear relationships and have been used in numerous studies to predict biomass, providing a significant advantage by combining multiple spectral and Remote Sens. 2024, 16, 3720 16 of 21 temporal variables, as seen in Bazzo [33]. The literature and current research suggest that combining spectral data with machine learning algorithms can significantly enhance biomass estimation accuracy, with each model offering specific advantages that can be exploited to improve estimation precision and robustness, serving as valuable tools for crop management and monitoring [73]. When applying different modeling approaches for biomass estimation via UAVs, it has been observed that each technique has its own strengths and limitations. For instance, linear regression is straightforward and easily interpretable but may fail to capture complex nonlinear relationships between input variables and biomass. Random Forest, which constructs multiple decision trees and averages their results, has proven robust and accurate but can be prone to overfitting with noisy datasets [32,91]. Support Vector Machines are effective in high-dimensional problems with small samples, although selecting appropriate parameters can be challenging. Artificial Neural Networks are powerful for modeling complex nonlinear relationships and have shown high accuracy in biomass estimation, though their interpretability and tendency to become stuck in local minima can be significant disadvantages [33]. These findings are consistent with other research utilizing UAVs for biomass estimation across different pasture types and vegetation [5], emphasizing the importance of selecting the appropriate model based on dataset characteristics and study objectives. 5. Conclusions The germination index showed low correlations with the target variable, rendering it insignificant for the predictive model. Statistical analyses revealed it is more closely associated with other variables not directly linked to the target variable. In contrast, dry matter is strongly correlated with NDVI values and plant height, indicating that these variables are more predictive. NDVI values, reflecting vegetation health along with dry matter and height, serve as better indicators of crop status and performance. Therefore, focusing on these variables enhances predictive model accuracy. Using UAVs to estimate forage biomass is a powerful tool that offers high resolution and flexibility in agricultural monitoring. However, to maximize its benefits, logistical, technical, and validation challenges must be overcome, alongside considerations of environ- mental conditions. With proper implementation, UAVs have the potential to revolutionize crop management practices and provide valuable data for agricultural decision-making. The analysis of predictive models, including Random Forest, Support Vector Ma- chines, and Neural Networks, has revealed significant variations in terms of accuracy and performance. The presentation of these results has been complemented by an expla- nation of the hyperparameters and the incorporation of visualizations, with the aim of providing a clearer and more practical perspective. This approach not only highlights the effectiveness of the models but also facilitates the interpretation and application of the results in real-world contexts, contributing to a more effective integration into agricultural decision-making. Consequently, an approach has been adopted that strives to make the models more accessible and useful. Author Contributions: J.U., S.P. (Samuel Pizarro), S.P. (Solanch Patricio) and D.F.: conceptualization; D.C. and I.C.: methodology; D.C. and S.P. (Samuel Pizarro): software; S.P.(Solanch Patricio) and D.F.: validation; J.P., L.E., D.C., K.O. and I.C.: formal analysis; D.C. and J.U.: investigation; J.U., D.C., K.O., I.C. and D.F.: resources; I.C. and D.C.: data curation; K.O.: writing—original draft preparation; K.O. and S.P. (Solanch Patricio): writing—review and editing, S.P (Samuel Pizarro). and L.E.: visualization; J.P. and S.P. (Samuel Pizarro): supervision; Z.O. and S.P. (Solanch Patricio): funding acquisition. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the project “Creación del servicio de agricultura de precisión en los Departamentos de Lambayeque, Huancavelica, Ucayali y San Martín 4 Departamentos” of the Ministry of Agrarian Development and Irrigation (MIDAGRI) of the Peruvian Government with grant number CUI 2449640. Remote Sens. 2024, 16, 3720 17 of 21 Data Availability Statement: The original contributions presented in the study are included in the article; further inquiries can be directed to the corresponding author. Acknowledgments: Special thanks are extended to the collaborators involved in field data collection and assistants of the Precision Agriculture Project (CUI 2449640) as well as other research programs of the “Estación Experimental Santa Ana”, INIA. Conflicts of Interest: The authors declare no conflicts of interest. References 1. Gupta, A.K.; Sharma, M.L.; Khan, M.A.; Pandey, P.K. Promotion of improved forage crop production technologies: Constraints and strategies with special reference to climate change. In Molecular Interventions for Developing Climate-Smart Crops: A Forage Perspective; Springer: Singapore, 2023; pp. 229–236. [CrossRef] 2. Trevisan, L.R.; Brichi, L.; Gomes, T.M.; Rossi, F. Estimating Black Oat Biomass Using Digital Surface Models and a Vegetation Index Derived from RGB-Based Aerial Images. Remote Sens. 2023, 15, 1363. [CrossRef] 3. Garland, G. Sustainable management of agricultural soils: Balancing multiple perspectives and tradeoffs. In EGU General Assembly Conference Abstracts; EGU: Vienna, Austria, 2023. [CrossRef] 4. Francaviglia, R.; Almagro, M.; Vicente-Vicente, J.L. Conservation Agriculture and Soil Organic Carbon: Principles, Processes, Practices and Policy Options. Soil Syst. 2023, 7, 17. [CrossRef] 5. Sharma, P.; Leigh, L.; Chang, J.; Maimaitijiang, M.; Caffé, M. Above-Ground Biomass Estimation in Oats Using UAV Remote Sensing and Machine Learning. Sensors 2022, 22, 601. [CrossRef] [PubMed] 6. Fodder, K.; Jimenez-Ballesta, R.; Srinivas Reddy, K.; Samuel, J.; Kumar Pankaj, P.; Gopala Krishna Reddy, A.; Rohit, J.; Reddy, K.S. Fodder Grass Strips for Soil Conservation and Soil Health. Chem. Proc. 2022, 10, 58. [CrossRef] 7. Katoch, R. Nutritional Quality of Important Forages. In Techniques in Forage Quality Analysis; Springer: Singapore, 2023; pp. 173–185. [CrossRef] 8. Barrett, B.A.; Faville, M.J.; Nichols, S.N.; Simpson, W.R.; Bryan, G.T.; Conner, A.J. Breaking through the feed barrier: Options for improving forage genetics. Anim. Prod. Sci. 2015, 55, 883–892. [CrossRef] 9. Kim, K.S.; Tinker, N.A.; Newell, M.A. Improvement of Oat as a Winter Forage Crop in the Southern United States. Crop Sci. 2014, 54, 1336–1346. [CrossRef] 10. McCartney, D.; Fraser, J.; Ohama, A. Annual cool season crops for grazing by beef cattle. A Canadian Review. Can. J. Anim. Sci. 2011, 88, 517–533. [CrossRef] 11. Kumar, S.; Vk, S.; Sanjay, K.; Priyanka; Gaurav, S.; Jyoti, K.; Kaushal, R. Identification of stable oat wild relatives among Avena species for seed and forage yield components using joint regression analysis. Ann. Plant Soil Res. 2022, 24, 601–605. [CrossRef] 12. Espinoza-Montes, F.; Nuñez-Rojas, W.; Ortiz-Guizado, I.; Choque-Quispe, D. Forage production and interspecific competition of oats (Avena sativa) and common vetch (Vicia sativa) association under dry land and high-altitude conditions. Rev. De. Investig. Vet. Del. Peru. 2018, 29, 1237–1248. [CrossRef] 13. INEI. Sistema Estadistico Nacional-Provincia de Lima 2018, 1–508. Available online: https://www.inei.gob.pe/media/ MenuRecursivo/publicaciones_digitales/Est/Lib1583/15ATOMO_01.pdf (accessed on 4 September 2024). 14. Aníbal, C.; Mayer, F. Producción de Carne y Leche Bovina en Sistemas Silvopastoriles; Instituto Nacional de Tecnología Agropecuaria: Buenos Aires, Argentina, 2017. 15. Santacoloma-Varón, L.E.; Granados-Moreno, J.E.; Aguirre-Forero, S.E. Evaluación de variables agronómicas, calidad del forraje y contenido de taninos condensados de la leguminosa Lotus corniculatus en respuesta a biofertilizante y fertilización química en condiciones agroecológicas de trópico alto andino colombiano. Entramado 2017, 13, 222–233. [CrossRef] 16. Mariana, M.P. Determinación de Variables Agronómicas del Cultivo de Maíz Mediante Imágenes Obtenidas Desde un Vehículo Aéreo no Tripulado (VANT). Thesis Instituto Mexicano de Tecnología del Agua, Jiutepec, Mexico, 2017. Available online: http://repositorio.imta.mx/handle/20.500.12013/1750 (accessed on 4 September 2024). 17. Watanabe, K.; Guo, W.; Arai, K.; Takanashi, H.; Kajiya-Kanegae, H.; Kobayashi, M.; Yano, K.; Tokunaga, T.; Fujiwara, T.; Tsutsumi, N.; et al. High-throughput phenotyping of sorghum plant height using an unmanned aerial vehicle and its application to genomic prediction modeling. Front. Plant Sci. 2017, 8, 254051. [CrossRef] [PubMed] 18. Matsuura, Y.; Heming, Z.; Nakao, K.; Qiong, C.; Firmansyah, I.; Kawai, S.; Yamaguchi, Y.; Maruyama, T.; Hayashi, H.; Nobuhara, H. High-precision plant height measurement by drone with RTK-GNSS and single camera for real-time processing. Sci. Rep. 2023, 13, 6329. [CrossRef] 19. Ji, Y.; Chen, Z.; Cheng, Q.; Liu, R.; Li, M.; Yan, X.; Li, G.; Wang, D.; Fu, L.; Ma, Y.; et al. Estimation of plant height and yield based on UAV imagery in faba bean (Vicia faba L.). Plant Methods 2022, 18, 26. [CrossRef] 20. Ibiev, G.Z.; Savoskina, O.A.; Chebanenko, S.I.; Beloshapkina, O.O.; Zavertkin, I.A. Unmanned Aerial Vehicles (UAVs)-One of the Digitalization and Effective Development Segments of Agricultural Production in Modern Conditions. In AIP Conference Proceedings; AIP Publishing: Melville, NY, USA, 2022; Volume 2661. [CrossRef] 21. Hütt, C.; Bolten, A.; Hüging, H.; Bareth, G. UAV LiDAR Metrics for Monitoring Crop Height, Biomass and Nitrogen Uptake: A Case Study on a Winter Wheat Field Trial. PFG-J. Photogramm. Remote Sens. Geoinf. Sci. 2023, 91, 65–76. [CrossRef] Remote Sens. 2024, 16, 3720 18 of 21 22. Plaza, J.; Criado, M.; Sánchez, N.; Pérez-Sánchez, R.; Palacios, C.; Charfolé, F. UAV Multispectral Imaging Potential to Monitor and Predict Agronomic Characteristics of Different Forage Associations. Agronomy 2021, 11, 1697. [CrossRef] 23. Radoglou-Grammatikis, P.; Sarigiannidis, P.; Lagkas, T.; Moscholios, I. A compilation of UAV applications for precision agriculture. Comput. Netw. 2020, 172, 107148. [CrossRef] 24. Lottes, P.; Khanna, R.; Pfeifer, J.; Siegwart, R.; Stachniss, C. UAV-Based Crop and Weed Classification for Smart Farming. In Proceedings of the 2017 IEEE International Conference on Robotics and Automation (ICRA), Singapore, 29 May–3 June 2017. 25. Munghemezulu, C.; Mashaba-Munghemezulu, Z.; Ratshiedana, P.E.; Economon, E.; Chirima, G.; Sibanda, S. Unmanned Aerial Vehicle (UAV) and Spectral Datasets in South Africa for Precision Agriculture. Data 2023, 8, 98. [CrossRef] 26. Gitelson, A.A.; Vina, A.; Arkebauer, T.J.; Rundquist, D.C.; Keydan, G.; Leavitt, B. Remote estimation of leaf area index and green leaf biomass in maize canopies. Geophys. Res. Lett. 2003, 30, 1248. [CrossRef] 27. Belton, D.; Helmholz, P.; Long, J.; Zerihun, A. Crop Height Monitoring Using a Consumer-Grade Camera and UAV Technology. PFG-J. Photogramm. Remote Sens. Geoinf. Sci. 2019, 87, 249–262. [CrossRef] 28. Hassan, M.A.; Yang, M.; Fu, L.; Rasheed, A.; Zheng, B.; Xia, X.; Xiao, Y.; He, Z. Accuracy assessment of plant height using an unmanned aerial vehicle for quantitative genomic analysis in bread wheat. Plant Methods 2019, 15, 37. [CrossRef] 29. Schreiber, L.V.; Atkinson Amorim, J.G.; Guimarães, L.; Motta Matos, D.; Maciel da Costa, C.; Parraga, A. Above-ground Biomass Wheat Estimation: Deep Learning with UAV-based RGB Images. Appl. Artif. Intell. 2022, 36. [CrossRef] 30. Bendig, J.; Bolten, A.; Bennertz, S.; Broscheit, J.; Eichfuss, S.; Bareth, G. Estimating Biomass of Barley Using Crop Surface Models (CSMs) Derived from UAV-Based RGB Imaging. Remote Sens. 2014, 6, 10395–10412. [CrossRef] 31. Jin, Y.; Yang, X.; Qiu, J.; Li, J.; Gao, T.; Wu, Q.; Zhao, F.; Ma, H.; Yu, H.; Xu, B. Remote Sensing-Based Biomass Estimation and Its Spatio-Temporal Variations in Temperate Grassland, Northern China. Remote Sens. 2014, 6, 1496–1513. [CrossRef] 32. Zhang, H.; Sun, Y.; Chang, L.; Qin, Y.; Chen, J.; Qin, Y.; Chen, J.; Qin, Y.; Du, J.; Yi, S.; et al. Estimation of Grassland Canopy Height and Aboveground Biomass at the Quadrat Scale Using Unmanned Aerial Vehicle. Remote Sens. 2018, 10, 851. [CrossRef] 33. Bazzo, C.O.G.; Kamali, B.; Hütt, C.; Bareth, G.; Gaiser, T. A Review of Estimation Methods for Aboveground Biomass in Grasslands Using, U.A.V. Remote Sens. 2023, 15, 639. [CrossRef] 34. Kurbanov, R.; Panarina, V.; Polukhin, A.; Lobachevsky, Y.; Zakharova, N.; Litvinov, M.; Rebouh, N.Y.; Kucher, D.E.; Gureeva, E.; Golovina, E.; et al. Evaluation of Field Germination of Soybean Breeding Crops Using Multispectral Data from UAV. Agronomy 2023, 13, 1348. [CrossRef] 35. Chen, R.; Chu, T.; Landivar, J.A.; Yang, C.; Maeda, M.M. Monitoring cotton (Gossypium hirsutum L.) germination using ultrahigh- resolution UAS images. Precis. Agric. 2018, 19, 161–177. [CrossRef] 36. Zhang, Y.; Liu, T.; He, J.; Yang, X.; Wang, L.; Guo, Y. Estimation of peanut seedling emergence rate of based on UAV visible light image. In Proceedings of the International Conference on Agri-Photonics and Smart Agricultural Sensing Technologies (ICASAST 2022), Zhengzhou, China, 4–6 August 2022; Volume 12349, pp. 259–265. [CrossRef] 37. Greaves, H.E.; Vierling, L.A.; Eitel, J.U.H.; Boelman, N.T.; Magney, T.S.; Prager, C.M.; Griffin, K.L. Estimating aboveground biomass and leaf area of low-stature Arctic shrubs with terrestrial LiDAR. Remote Sens. Environ. 2015, 164, 26–35. [CrossRef] 38. Li, K.Y.; de Lima, R.S.; Burnside, N.G.; Vahtmäe, E.; Kutser, T.; Sepp, K.; Pinheiro, V.H.C.; Yang, M.-D.; Vain, A.; Sepp, K. Toward Automated Machine Learning-Based Hyperspectral Image Analysis in Crop Yield and Biomass Estimation. Remote Sens. 2022, 14, 1114. [CrossRef] 39. Cáceres, Y.Z.; Torres, B.C.; Archi, G.C.; Zanabria Mallqui, R.; Pinedo, L.E.; Trucios, D.C.; Ortega Quispe, K.A. Analysis of Soil Quality through Aerial Biomass Contribution of Three Forest Species in Relict High Andean Forests of Peru. Malaysian Journal Soil Science. 2024, 28, 38–52. 40. Naveed Tahir, M.; Zaigham Abbas Naqvi, S.; Lan, Y.; Zhang, Y.; Wang, Y.; Afzal, M.; Cheema, M.J.M.; Amir, S. Real time estimation of chlorophyll content based on vegetation indices derived from multispectral UAV in the kinnow orchard. Int. J. Precis. Agric. Aviat. 2018, 1, 24–31. [CrossRef] 41. Fei, S.; Hassan, M.A.; Xiao, Y.; Su, X.; Chen, Z.; Cheng, Q.; Duan, F.; Chen, R.; Ma, Y. UAV-based multi-sensor data fusion and machine learning algorithm for yield prediction in wheat. Precis. Agric. 2023, 24, 187–212. [CrossRef] 42. Yue, J.; Yang, G.; Tian, Q.; Feng, H.; Xu, K.; Zhou, C. Estimate of winter-wheat above-ground biomass based on UAV ultrahigh- ground-resolution image textures and vegetation indices. ISPRS J. Photogramm. Remote Sens. 2019, 150, 226–244. [CrossRef] 43. Zhai, W.; Li, C.; Cheng, Q.; Mao, B.; Li, Z.; Li, Y.; Ding, F.; Qin, S.; Fei, S.; Chen, Z. Enhancing Wheat Above-Ground Biomass Esti- mation Using UAV RGB Images and Machine Learning: Multi-Feature Combinations, Flight Height, and Algorithm Implications. Remote Sens. 2023, 15, 3653. [CrossRef] 44. Quirós, J.J.; McGee, R.J.; Vandemark, G.J.; Romanelli, T.; Sankaran, S. Field phenotyping using multispectral imaging in pea (Pisum sativum L) and chickpea (Cicer arietinum L). Eng. Agric. Environ. Food 2019, 12, 404–413. [CrossRef] 45. Ortega Quispe, K.A.; Valerio Deudor, L.L. Captación y almacenamiento pluvial como modelo histórico para conservación del agua en los Andes peruanos. Desafios 2023, 14, e385. [CrossRef] 46. IGP. Atlas Climático de Precipitación y Temperatura del Aire de la Cuenca del Río Mantaro; Consejo Nacional del Ambiente: Lima, Peru, 2005. 47. Ccopi-Trucios, D.; Barzola-Rojas, B.; Ruiz-Soto, S.; Gabriel-Campos, E.; Ortega-Quispe, K.; Cordova-Buiza, F. River Flood Risk Assessment in Communities of the Peruvian Andes: A Semiquantitative Application for Disaster Prevention. Sustainability 2023, 15, 13768. [CrossRef] Remote Sens. 2024, 16, 3720 19 of 21 48. ISTA. Reglas Internacionales para el Análisis de las Semillas; International Seed Testing Association: Bassersdorf, Switzerland, 2016; pp. 1–384. [CrossRef] 49. Hijmans, J. Package “terra” Spatial Data Analysis 2024. Available online: https://cran.r-project.org/web/packages/terra/terra. pdf (accessed on 4 September 2024). 50. Huete, A.R.; Liu, H.Q.; Batchily, K.; Van Leeuwen, W. A comparison of vegetation indices over a global set of TM images for EOS-MODIS. Remote Sens. Environ. 1997, 59, 440–451. [CrossRef] 51. Koppe, W.; Li, F.; Gnyp, M.L.; Miao, Y.; Jia, L.; Chen, X.; Zhang, F.; Bareth, G. Evaluating multispectral and hyperspectral satellite remote sensing data for estimating winter wheat growth parameters at regional scale in the North China plain. Photogramm. Fernerkund. Geoinf. 2010, 2010, 167–178. [CrossRef] 52. Gitelson, A.A.; Kaufman, Y.J.; Merzlyak, M.N. Use of a green channel in remote sensing of global vegetation from EOS-MODIS. Remote Sens. Environ. 1996, 58, 289–298. [CrossRef] 53. Strong, C.J.; Burnside, N.G.; Llewellyn, D. The potential of small-Unmanned Aircraft Systems for the rapid detection of threatened unimproved grassland communities using an Enhanced Normalized Difference Vegetation Index. PLoS ONE 2017, 12, e0186193. [CrossRef] [PubMed] 54. Jordan, C.F. Derivation of Leaf-Area Index from Quality of Light on the Forest Floor. Ecology 1969, 50, 663–666. [CrossRef] 55. Huete, A.; Didan, K.; Miura, T.; Rodriguez, E.P.; Gao, X.; Ferreira, L.G. Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sens. Environ. 2002, 83, 195–213. [CrossRef] 56. Wu, C.; Niu, Z.; Gao, S. The potential of the satellite derived green chlorophyll index for estimating midday light use efficiency in maize, coniferous forest and grassland. Ecol. Indic. 2012, 14, 66–73. [CrossRef] 57. Gitelson, A.A.; Gritz, Y.; Merzlyak, M.N. Relationships between leaf chlorophyll content and spectral reflectance and algorithms for non-destructive chlorophyll assessment in higher plant leaves. J. Plant Physiol. 2003, 160, 271–282. [CrossRef] 58. Gitelson, A.A.; Kaufman, Y.J.; Stark, R.; Rundquist, D. Novel algorithms for remote estimation of vegetation fraction. Remote Sens. Environ. 2002, 80, 76–87. [CrossRef] 59. Richardson, A.J.; Wiegand, C.L. Distinguishing Vegetation from Soil Background Information. Photogramm. Eng. Remote Sens. 1977, 43, 1541–1552. 60. Vincini, M.; Frazzi, E.; D’Alessio, P. A broad-band leaf chlorophyll vegetation index at the canopy scale. Precis. Agric. 2008, 9, 303–319. [CrossRef] 61. Schleicher, T.D.; Bausch, W.C.; Delgado, J.A.; Ayers, P.D. Evaluation and Refinement of the Nitrogen Reflectance Index (NRI) for Site-Specific Fertilizer Management. In Proceedings of the 2001 ASAE Annual Meeting, Sacramento, CA, USA, 29 July–1 August 2001; p. 1. [CrossRef] 62. Daughtry, C.S.T.; Walthall, C.L.; Kim, M.S.; De Colstoun, E.B.; McMurtrey, J.E. Estimating Corn Leaf Chlorophyll Concentration from Leaf and Canopy Reflectance. Remote Sens. Environ. 2000, 74, 229–239. [CrossRef] 63. Eitel, J.U.H.; Long, D.S.; Gessler, P.E.; Smith, A.M.S. Using in-situ measurements to evaluate the new RapidEyeTM satellite series for prediction of wheat nitrogen status. Int. J. Remote Sens. 2007, 28, 4183–4190. [CrossRef] 64. Pereira, F.R.d.S.; de Lima, J.P.; Freitas, R.G.; Dos Reis, A.A.; do Amaral, L.R.; Figueiredo, G.K.D.A.; Lamparelli, R.A.; Magalhães, P.S.G. Nitrogen variability assessment of pasture fields under an integrated crop-livestock system using UAV, PlanetScope, and Sentinel-2 data. Comput. Electron. Agric. 2022, 193, 106645. [CrossRef] 65. Karnati, R.; Prasad, M.V.S. a Prediction of Crop Monitoring Indices (NDVI,MSAVI,RECI) and Estimation of Nitrogen Concentra- tion on Leaves for Possible of Optimizing the Time of Harvest with the Help of Sensor Networks in Guntur Region, Andhra Pradesh, India. with agent based modeling. Int. J. Adv. Sci. Comput. Appl. 2023, 2, 19–30. [CrossRef] 66. Kureel, N.; Sarup, J.; Matin, S.; Goswami, S.; Kureel, K. Modelling vegetation health and stress using hypersepctral remote sensing data. Model. Earth Syst. Environ. 2022, 8, 733–748. [CrossRef] 67. Yang, C.; Everitt, J.H.; Bradford, J.M.; Murden, D. Airborne hyperspectral imagery and yield monitor data for mapping cotton yield variability. Precis. Agric. 2004, 5, 445–461. [CrossRef] 68. Yang, K.; Tu, J.; Chen, T. Homoscedasticity: An overlooked critical assumption for linear regression. Gen. Psychiatr. 2019, 32, e100148. [CrossRef] 69. Hope, T.M.H. Linear regression. In Machine Learning: Methods and Applications to Brain Disorders; Acamedic Press: Cambridge, MA, USA, 2020; pp. 67–81. [CrossRef] 70. Dhulipala, S.; Patil, G.R. Freight production of agricultural commodities in India using multiple linear regression and generalized additive modelling. Transp. Policy 2020, 97, 245–258. [CrossRef] 71. Hastie, T.; Tibshirani, R.; Friedman, J. Springer Series in Statistics The Elements of Statistical Learning Data Mining, Inference, and Prediction. Math. Intell. 2008, 27, 83–85. 72. Bhavsar, H.; Ganatra, A. A Comparative Study of Training Algorithms for Supervised Machine Learning. Int. J. Soft Comput. Eng. (IJSCE) 2012, 2, 2231–2307. 73. Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [CrossRef] 74. Liaw, A.; Wiener, M. Classification and Regression by randomForest. R News 2002, 2, 18–22. 75. Mammone, A.; Turchi, M.; Cristianini, N. Support vector machines. Wiley Interdiscip. Rev. Comput. Stat. 2009, 1, 283–289. [CrossRef] Remote Sens. 2024, 16, 3720 20 of 21 76. Capparuccia, R.; De Leone, R.; Marchitto, E. Integrating support vector machines and neural networks. Neural Netw. 2007, 20, 590–597. [CrossRef] [PubMed] 77. Lecun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature 2015, 521, 436–444. [CrossRef] [PubMed] 78. Leng, Y.; Xu, X.; Qi, G. Combining active learning and semi-supervised learning to construct SVM classifier. Knowl. Based Syst. 2013, 44, 121–131. [CrossRef] 79. Williams, H.A.M.; Jones, M.H.; Nejati, M.; Seabright, M.J.; Bell, J.; Penhall, N.D.; Barnett, J.J.; Duke, M.D.; Scarfe, A.J.; Ahn, H.S.; et al. Robotic kiwifruit harvesting using machine vision, convolutional neural networks, and robotic arms. Biosyst. Eng. 2019, 181, 140–156. [CrossRef] 80. Cortes, C.; Vapnik, V.; Saitta, L. Support-Vector Networks Editor; Kluwer Academic Publishers: New York, USA, 1995; Volume 20. 81. Przybyło, J.; Jabłoński, M. Using Deep Convolutional Neural Network for oak acorn viability recognition based on color images of their sections. Comput. Electron. Agric. 2019, 156, 490–499. [CrossRef] 82. A Ilemobayo, J.; Durodola, O.; Alade, O.; J Awotunde, O.; T Olanrewaju, A.; Falana, O.; Ogungbire, A.; Osinuga, A.; Ogunbiyi, D.; Ifeanyi, A.; et al. Hyperparameter Tuning in Machine Learning: A Comprehensive Review. J. Eng. Res. Rep. 2024, 26, 388–395. [CrossRef] 83. Defalque, G.; Santos, R.; Bungenstab, D.; Echeverria, D.; Dias, A.; Defalque, C. Machine learning models for dry matter and biomass estimates on cattle grazing systems. Comput. Electron. Agric. 2024, 216, 108520. [CrossRef] 84. Natekin, A.; Knoll, A. Gradient boosting machines, a tutorial. Front. Neurorobot 2013, 7, 21. [CrossRef] 85. Hornik, K. Resampling Methods in R: The boot Package. R News 2002, 2, 2–7. 86. Wei, T.; Simko, V.; Levy, M.; Xie, Y.; Jin, Y.; Zemla, J. Package “Corrplot” Title Visualization of a Correlation Matrix Needs Compilation. 2022. Available online: https://cran.r-project.org/web/packages/corrplot/corrplot.pdf (accessed on 4 September 2024). 87. Greenacre, M.; Groenen, P.J.F.; Hastie, T.; D’Enza, A.I.; Markos, A.; Tuzhilina, E. Principal component analysis. Nat. Rev. Methods Primers 2022, 2, 100. [CrossRef] 88. Gilbertson, J.K.; van Niekerk, A. Value of dimensionality reduction for crop differentiation with multi-temporal imagery and machine learning. Comput. Electron. Agric. 2017, 142, 50–58. [CrossRef] 89. Jollife, I.T.; Cadima, J. Principal component analysis: A review and recent developments. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2016, 374, 20150202. [CrossRef] 90. Hasan, B.M.S.; Abdulazeez, A.M. A Review of Principal Component Analysis Algorithm for Dimensionality Reduction. J. Soft Comput. Data Min. 2021, 2, 20–30. [CrossRef] 91. Lussem, U.; Bolten, A.; Menne, J.; Gnyp, M.L.; Schellberg, J.; Bareth, G. Estimating biomass in temperate grassland with high resolution canopy surface models from UAV-based RGB images and vegetation indices. J. Appl. Remote Sens. 2019, 13, 1. [CrossRef] 92. Coelho, A.P.; de Faria, R.T.; Leal, F.T.; Barbosa, J.d.A.; Dalri, A.B.; Rosalen, D.L. Estimation of irrigated oats yield using spectral indices. Agric. Water Manag. 2019, 223, 105700. [CrossRef] 93. Chai, T.; Draxler, R.R. Root mean square error (RMSE) or mean absolute error (MAE)? -Arguments against avoiding RMSE in the literature. Geosci. Model. Dev. 2014, 7, 1247–1250. [CrossRef] 94. Cort, W.J.; Matsuura, K. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Clim. Res. 2005, 30, 79–82. 95. Hyndman, R.J.; Koehler, A.B. Another look at measures of forecast accuracy. Int. J. Forecast. 2006, 22, 679–688. [CrossRef] 96. Peters, K.C.; Hughes, M.P.; Daley, O. Field-scale calibration of the PAR Ceptometer and FieldScout CM for real-time estimation of herbage mass and nutritive value of rotationally grazed tropical pasture. Smart Agric. Technol. 2022, 2, 100037. [CrossRef] 97. Lu, B.; Proctor, C.; He, Y. Investigating different versions of PROSPECT and PROSAIL for estimating spectral and biophysical properties of photosynthetic and non-photosynthetic vegetation in mixed grasslands. GIsci Remote Sens. 2021, 58, 354–371. [CrossRef] 98. Zhang, Q.; Xu, L.; Zhang, M.; Wang, Z.; Gu, Z.; Wu, Y.; Shi, Y.; Lu, Z. Uncertainty analysis of remote sensing pretreatment for biomass estimation on Landsat OLI and Landsat ETM+. ISPRS Int. J. Geoinf. 2020, 9, 48. [CrossRef] 99. Lu, D.; Chen, Q.; Wang, G.; Liu, L.; Li, G.; Moran, E. A survey of remote sensing-based aboveground biomass estimation methods in forest ecosystems. Int. J. Digit. Earth 2016, 9, 63–105. [CrossRef] 100. Song, Q.; Albrecht, C.M.; Xiong, Z.; Zhu, X.X. Biomass Estimation and Uncertainty Quantification From Tree Height. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2023, 16, 4833–4845. [CrossRef] 101. Grüner, E.; Astor, T.; Wachendorf, M. Biomass prediction of heterogeneous temperate grasslands using an SFM approach based on UAV imaging. Agronomy 2019, 9, 54. [CrossRef] 102. Barboza, T.O.C.; Ardigueri, M.; Souza, G.F.C.; Ferraz, M.A.J.; Gaudencio, J.R.F.; Santos, A.F.d. Performance of Vegetation Indices to Estimate Green Biomass Accumulation in Common Bean. AgriEngineering 2023, 5, 840–854. [CrossRef] Remote Sens. 2024, 16, 3720 21 of 21 103. Pizarro, S.; Pricope, N.G.; Figueroa, D.; Carbajal, C.; Quispe, M.; Vera, J.; Alejandro, L.; Achallma, L.; Gonzalez, I.; Salazar, W.; et al. Implementing Cloud Computing for the Digital Mapping of Agricultural Soil Properties from High Resolution UAV Multispectral Imagery. Remote Sens. 2023, 15, 3203. [CrossRef] 104. Sinde-González, I.; Gil-Docampo, M.; Arza-García, M.; Grefa-Sánchez, J.; Yánez-Simba, D.; Pérez-Guerrero, P.; Abril-Porras, V. Biomass estimation of pasture plots with multitemporal UAV-based photogrammetric surveys. Int. J. Appl. Earth Obs. Geoinf. 2021, 101, 102355. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.