Extraction of greengage planting information from multi-source remote sensing images based on GEE
The planting area of greengage in eastern Guangdong is mostly a complex terrain with both mountains and hills.It is difficult to measure the planting area and spatial distribution of greengage in the field.This study focused on Shanwei City and Jieyang City,and used the random forest algorithm to explore the performance of four different feature combinations in the classification accuracy of greengage based on Sentinel-1 and Sentinel-2 multi-spectral remote sensing image data.In addition,the accuracy of Sentinel-1 and Sentinel-2 images in spatial mapping was analyzed in depth.The results showed that the optimal classification accuracy of greengage was obtained when Sentinel-1 and Sentinel-2 images were processed by combination 4(spectral feature+vegetation index+VV/VH+texture feature)and random forest algo-rithm.The overall accuracy,Kappa coefficient and mapping accuracy were 96.55％,0.957 6 and 98.03％,respectively.Compared with the official statistics,the ac-curacy of estimating the planting area of greengage by u-sing combination 4 in Puning City of Jieyang City and Luhe County of Shanwei City was 99.70％and 99.20％,respec-tively.The comprehensive utilization of Sentinel-1 and Sentinel-2 multi-source remote sensing data and the classification methods using multiple features could accurately identify the planting area of greengage and improve the mapping accuracy.The results of this study not only provide accurate planting area estimation and spatial distribution information for greengage growers in eastern Guangdong,but also provide technical reference for the formulation of greengage planting management and pest control strategies.
万方数据
维普
