Research of Image Change Detection Method Based on Spatial Autocorrelation Statistics

It is difficult for the existing image change detection algorithms to achieve fast change detection, due to their complexities and large calculated quantities. It applied the spatial autocorrelation statistics method in the image change detection, and identified potential changes in the image area by comparing the spatial autocorrelation coefficients of images acquired at different times. With this method utilized in the change detection of the coral reef ecosystem, it detected the changes of two scene TM images. The results have validated the simplicity and accuracy of this method.