Research on crop identification method based on two-phase classification using remote sensing in large scale

The paddy rice is one of the major crops in China and its planting areas increase year by year in Northeast China. In this study, a crop identification system in large scale based on two- phase classification using Remote Sensing (RS) is designed to do the rice identification. It includes the stratified sampling method to select the sample area using RS, and different spatial resolution RS images. To improve the accuracy of identification and classification and reduce the error of sampling and extrapolating, this study takes Jilin province of China as the experimental area, collects the ground survey data and makes assumptions that the large scale and classification accuracy of the above method can meet the requirement of production. And, Quickbird shows that the overall accuracy achieved is 96%. Therefore this research provides a technical support for the RS monitoring system of the Ministry of Agriculture of China.