草业科学2026，Vol.43Issue(3)：719-730,12.DOI:10.11829/j.issn.1001-0629.2025-0240
基于无人机遥感的企业苜蓿产量估测研究
Yield estimation of enterprise alfalfa based on unmanned aerial vehicle multispectral imaging
摘要
Abstract
Alfalfa(Medicago sativa)occupies an important position in global animal husbandry due to its high yield and rich nutritional value.With the rapid development of agricultural information technology,the innovative application of unmanned aerial vehicle(UAV)remote sensing technology provides a new multifunctional technical means for enterprises to monitor the growth dynamics of alfalfa.This study designed irrigation,fertilization,and water-fertilizer interaction experiments at the production base of the enterprise.UAV multispectral data were collected for three growth periods,extracting 14 vegetation indices(VI)and 40 texture features(TF).Multiple linear regression(MLR),random forest(RF),and support vector machine(SVM)models were used to construct alfalfa yield estimation models.The research findings indicated that:1)The average canopy reflectance of alfalfa was highest under a gradient of medium-high water and high fertilizer.2)The normalized difference red edge index,vegetation stress index,and mean values of green-and red-light bands exhibited the most significant correlation with yield.The synergistic effects of water and fertilizer effectively enhanced the correlation between VI,TF,and yield(P<0.05).3)The combination of VI and TF achieved the best estimation results in a RF model,with an R2 of 0.89,low error,and no overfitting.关键词
紫花苜蓿/无人机/多光谱/产量/机器学习/植被指数/纹理特征Key words
alfalfa/unmanned aerial vehicle/multispectral/yield/machine learning/vegetation index/texture feature引用本文复制引用
林泽云,李文龙,赵玲,李思清,刘星园,梁天刚..基于无人机遥感的企业苜蓿产量估测研究[J].草业科学,2026,43(3):719-730,12.基金项目
农业农村部科学技术司国家牧草产业技术体系遥感监测与智能管理岗位科学家项目(CARS34) （CARS34）
