Crop planting region extraction based on cloud model and multi-temporal MODIS data

Crop growth monitoring is an important application of remote sensing. However, the crop cultivation area cannot be easily distinguished from other areas in the remote sensing data because of the various types of land cover. Therefore, the crop cultivation area should be extracted before remote sensing data can be used. In this paper, a new method is presented to extract wheat planting region. Two MODIS images representing different times were adopted according to the characteristics of the growth of wheat. The training sets selected in two MODIS images were calculated by backward cloud algorithm to generate the digital features which reflect two disparate qualitative concepts. Next, the membership grade of the NDVI value of each pixel in two MODIS images to the corresponding qualitative concept was computed. Finally, the calculation results of two images were overlapped to extract wheat planting region. Compared with other current methods of crop planting extraction based on color composition and classification, this method excels in higher accuracy, unlimited amounts of images and low workload of choosing training sets which only need to sample the crop region. Therefore, this method is suitable for large-scale crop planting region extraction.