Skip to Main Content
Several land cover change detection approaches have been developed such as traditional post-classification cross-tabulation, cross-correlation analysis, neural networks and knowledge-based expert systems. But the procedures for analyzing remote sensing image have relied upon pixel-based methods. So, these methods are restricted by the spatial resolution of remote sensing image and widely used with moderate resolution image. With the advent of high spatial resolution remote sensing data, this situation is different. The characteristics of high spatial resolution remote sensing image are different essentially with these of moderate resolution image. So, these traditional methods are not applicable with the new data. In order to solve the problems and make full use of the advantages of high spatial resolution data, an object-oriented land cover change detection approach was developed in this paper. The whole image was segmented into several objects by texture characteristics and relationship between each pixel. Then, a supervised classification was carried out at the object level. The change detection was also realized by the comparison between images of different times.