Winter wheat area extraction and estimation based on MODIS-VI time series and multi-temporal HJ CCD images in Jiangsu province, China

摘要

This paper proposed a winter wheat extraction and estimation model for Jiangsu province, China, using HJ multi temporal images and MODIS-VI time series. The illustrations of winter wheat's representative phenological calendar features presented on EVI series are taken as optimal indicators. Combined with field surveys, HJ EVI and NDVI features are depicted, and winter wheat area on typical research area are obtained and verified. Several EVI scenes which behaved critical phenological features are chosen, and then a quantitative function between the MODIS EVI and crop planted area percentage is constructed. Winter wheat extraction result from HJ images is used for the sample selection and coefficients calculation. The winter wheat for the whole province is extracted and calculated using the model. Estimation precision is evaluated by the official statistic data collected. The improved high precision shows the effectiveness of the model and confirms the advantage of the two image types' combination.