Automatic Change Detection and Updating of Topographic Databases by using Level Set Active Contours: Application to Lake Chang

In order to keep up-to-date geospatial data in topographic databases, automatic change detection and data updating is required. In the presented paper, we investigate the automatic change detection of geospatial data by using level set active contours. We propose an approach that is based on region comparison between two multi-temporal datasets. Firstly, the regions are extracted from two co-registered images taken apart in time by using level set based active contours segmentation. Then, the change detection is performed by comparing spatially the resulting region segments from the two images. The approach is validated by experiments relating to the change detection of lake surfaces by using Lansat7 multi-spectral imagery.