Indexing an image database by shape content using curvature scale space

We introduce a very fast and reliable algorithm for shape similarity retrieval in large image databases which is robust with respect to noise, scale and orientation changes of the objects. The curvature scale space image is used to represent the shapes of object boundary contours. Since the algorithm uses the global information of boundaries of objects, it is sensitive to major occlusion, but some minor occlusion can be detected by the algorithm. We tested our method on a database of 450 images of marine animals with a vast variety of shapes, with very good results which we present in this paper