Abstract: The acquisition of large area paddy rice acreage by remote sensing is one of the chief fields in the application of remote sensing in agriculture. And multi-resolution remotely sensed data method plays an important role. In this study, paddy rice result, acquired from SPOT data, severs as high resolution samples, while MODIS-EVI time series data serves as low resolution data, and then using support vector machine (SVM) method obtain the paddy rice coverage. Two important advancements have been found from this study: 1) The similar index which is based on the MODIS-EVI time series RS-data and filed work samples can describe the paddy rice coverage obviously; 2) The method of choosing sample based on the image, similar index image, comparability has firmly theory foundation which means when the image comparability higher the accuracy better. While at the same, the method’s stability has validated by repetitious tests at the different situations of entire similar samples, the result shows that the method has better stability and can serves as a substitution in acquisition of large area paddy rice coverage.

Key words: plant area, paddy rice, similar index analysis, support vector machine