Study on information extraction of rape acreage based on TM remote sensing image

China's rape acreage and total output of rapeseeds ranks among the top in the world, accounting for more than 30% of the world's total rape acreage and output of rapeseeds. This paper takes Landsat TM as the main data source in conducting the study of extracting rape acreage information in the Shou County, Anhui Province. Through analysis and calculations of phenological diversity, spectral discrimination, etc. of various main vegetations, with remote sensing image in each growth stages of rape within the studied area, this paper concludes that the optimal time period for information extraction of rape acreage based on TM image is the flowering period for rape. This paper adopts confusion matrix calculation to compare non- supervised and supervised classification methods in extracting rape acreage information using remote sensing image. The results show that the classification results of Mahalanobis Distance method and Isodate non-supervised classification method yielded relatively good results. In which, the Isodate non-supervised classification method combined with human visual inspection can extract the rape planting area information with higher precision and efficiency. The study shows that the method by utilizing TM remote sensing data to extract information of rape acreage can get a relatively satisfactory result. We believe the rape acreage remote sensing identification technology can provide a scientific reference to the understanding of China's rape planting situation.