Extract agricultural land use information from remote sensing image based on object-oriented classification algorithm

On the research of agricultural Non-point source pollution, agricultural land use data is the most important basic information, it's hard to obtain by statistical means and it update frequently. By the help of remote sensing image and information extraction techniques, agricultural land use information can be extracted with low-cost and fast and accurately. In this study, we established index system of agricultural land use classification in line with agricultural non-point source pollution, and use the Japanese ALOS satellite as a data source, extract agricultural land use information of Chongming Island by object-oriented remote sensing image classification algorithm. After a wild field verification, classification accuracy can meet agricultural nonpoint source pollution requirements.