Topology based corner detection

Corner detection is one of the most important tasks in the field of image processing. It is one of basic operations used in higher level tasks such as obect recognition, face recognition and segmentation. So far, many algorithms are proposed for corner detection. Among them, the most widely used algorithms are SUSAN, CSS, Kitchen Rosenfeld and Plessey algorithms due to their successes and requiring less computation power. In this thesis work, a fuzzy topological corner detection algorithm is presented. The proposed method enhances the algorithm by decreasing noise effect, thus ensures lower false detections and more true corners. This thesis work is based on extending the SUSAN algorithm with the proposed method. SUSAN algorithm is chosen because it is generally more successful than the others. According to the results obtained, topological version is superior to the conventional version but it is slightly slower. Especially, false corner detection rates are reduced by %40.