基于冠层高光谱遥感对加工番茄产量的估算模型

Estimation Models of Processing Tomato Yield Based on Canopy Hyperspectral Remote Sensing

[目的]对加工番茄的产量进行遥感估测。[方法]以ASD FieldSpec光谱仪实测大田中不同生育期加工番茄的冠层高光谱及其产量,采用单时相线性逐步回归和复合回归,首次建立了加工番茄高光谱与产量的估算模型。[结果]在坐果期光谱参量与产量相关性最大,而其他时期的光谱参量与产量相关性均达到了显著水平(P<0.05);多时相复合回归模型以4个生育期与产量的复合回归最为理想。[结论]利用高光谱遥感来监测加工番茄的生长状况,可以最终对加工番茄的产量进行遥感估测。

[ Objective] The aim was to estimate the yield of processing tomato with remote sensing. [ Method] Canopy reflectance of processing tomato was measured with ASD FieldSpec during different stages. And the yield of processing tomato was collected. The relationship between canopy reflectance and processing tomato yield analyzed with single phase linear stepwise regression and multivariate regression for the first time. [ Result] Hyperspectral parameters in fruit-set period had better correlation to yields. While hyperspectral parameters in other periods had a good relation to yields in different growing stages and correlation was extremely remarkable （ P 〈 0.05）. By comparison, four periods stepwise regression model is the best method for estimate processing tomato yield with remote sensing data. [Conclusion] The growth status of processing tomato can be monitored with hyperspectral remote sensing. Fitrolly the yield of processing tomato can be estimated.