基于無人機(jī)高光譜和集成學(xué)習(xí)的春小麥葉綠素含量反演

打開文本圖片集

中圖分類號：S127 文獻(xiàn)標(biāo)志碼：A 文章編號：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

葉綠素含量（leaf chlorophyll content，LCC）作為表征植物生長過程的關(guān)鍵生理參數(shù)，既是評價(jià)植物當(dāng)前營養(yǎng)狀況和生長發(fā)育水平的有效指標(biāo)，也是衡量植物長勢的指示器，LCC變化直接反映作物脅迫狀況、光合作用能力和衰老進(jìn)程的信息[2-3]。（剩余18423字）