基于无人机高光谱和集成学习的春小麦叶绿素含量反演

打开文本图片集
中图分类号:S127 文献标志码:A 文章编号:1008-0864(2025)06-0093-11
Chlorophyll Content Inversion of Spring Wheat Based on Unmanned Aerial Vehicle Hyperspectral and Integrated Learning
HU Sile,BAO Yulong*,Tubuxinbayaer,TAO Jifeng,GUO Enliang (KeyLaboratoryofGeographicResearchontheMongolianPlateauinInnerMongoliaAwtonomousRegion,ollgeofGeogaphical Science,InnerMongoliaNormalUniversity,HohhotO1oO22,China)
Abstract:Chlorophyllcontentisakey indicator for monitoringcropgrowth,and itsrapid,effctiveandaccurate estimation is crucial for assessing crop health.Bycollecting unmanned aerial vehicle(UAV) hyperspectral images from3growth stagesand combining themwith ground-measured chlorophylldata,variousmachine learningand ensemblelearning models were employed to estimate the chlorophyll content in spring wheat,and the estimation accuracyof different modelswere compared.The results showedthat the canopyreflectance of spring wheat was generallyconsistent acrossdifferent growth stages,but significant diferences in spectralreflectance intensity were observedin the770\~9OO nmwavelength range.16spectral indices allshowed significantcorelationswithchlorophyll content,among which optimized vegetationindex1,plant biochemical indexand normalized diferencerededge indexmaintained high correlation throughout theentire growth cycle.Thepredictionaccuracyof the Stackingand Voting ensemble learning modelswas higher than thebasic models,withthe Voting ensemble model performing particularly well.In the test set,determination coefficient( R2 )values of 3 growth stages were 0.78,0.77and0.73, and root mean square error(RMSE)values were 8.70,11.36 and 16.17,respectively. Compared with random forest, support vector regression,K_nearest neighbor and ridge regression models,the R2 of the Voting model was on average0.17,0.14and0.22 higher,andtheRMSE was 4.64,2.54and6.51lower,indicating itssuperiorpredictive ability.Above results provided new perspectives and methods for precision agriculture and crop health monitoring. Key words:unmanned aerial vehicle(UAV);hyperspectral remotesensing;spring wheat;chlorophyll content; ensemble learning
叶绿素含量(leaf chlorophyll content,LCC)作为表征植物生长过程的关键生理参数,既是评价植物当前营养状况和生长发育水平的有效指标,也是衡量植物长势的指示器,LCC变化直接反映作物胁迫状况、光合作用能力和衰老进程的信息[2-3]。(剩余18423字)