Paddy Fields Area Extraction Based on Landsat Data with Integrating Temporal Features
Taking Nan County, a typical paddy production county in Dongting Lake area, as the study area, we used remote sensing data to derive the paddy area. According to the return cycle of the Landsat and rainy weather of the middle and lower reaches of Yangtze River, we utilized spatio-temporal adaptive reflectance fusion model(STARFM) to fuse MODIS data with high temporal resolution and Landsat data with moderate spatial resolution, and produced the time series Landsat TM data with temporal features. And then, we used Landsat data with integrating temporal features and SVM to extract the paddy fields area based on crop phenology characteristics. The results indicate that the classification accuracy of paddy has been improved by using time series Landsat NDVI data.
【CateGory Index】： P237