Rotation and Scale Invariant Shape Description Using the Contour Segment Curvature

This paper presents a shape description method based on contour segment curvature (CSC). The CSC is defined as the ratio of the line length connecting two endpoints of a contour segment to its curve length. To extract consistent contour segment, the concept of overlapped contour segment is introduced. The rotation and scale invariant CSC can be extracted through the use of the overlapped contour segment. The proposed method describes the shape of objects with feature vectors that represents the distribution of the CSC, and measures the similarity by comparing the feature vector acquired from the corresponding unit-length segment. The experimental results show that the proposed method is not only invariant to rotation and scale but also superior to the NCCH and the TRP method in clustering power. Furthermore, the performance improvement is expected by adding the distance information to the CSC.

[1]  Sven Loncaric,et al.  A survey of shape analysis techniques , 1998, Pattern Recognit..

[2]  Arnold W. M. Smeulders,et al.  Image Databases and Multi-Media Search , 1998, Image Databases and Multi-Media Search.

[3]  Daniel Tretter,et al.  A Web-Based Secure System for the Distributed Printing of Documents and Images , 1998, J. Vis. Commun. Image Represent..

[4]  Guojun Lu,et al.  Review of shape representation and description techniques , 2004, Pattern Recognit..

[5]  Shih-Fu Chang,et al.  Image Retrieval: Current Techniques, Promising Directions, and Open Issues , 1999, J. Vis. Commun. Image Represent..

[6]  Chin-Chen Chang,et al.  A shape recognition scheme based on relative distances of feature points from the centroid , 1991, Pattern Recognition.

[7]  Ari Visa,et al.  Shape recognition of irregular objects , 1996, Other Conferences.

[8]  Thomas S. Huang,et al.  A Modified Fourier Descriptor for Shape Matching in MARS , 1998, Image Databases and Multi-Media Search.

[9]  Cyrus Shahabi,et al.  Image retrieval by shape: a comparative study , 2000, 2000 IEEE International Conference on Multimedia and Expo. ICME2000. Proceedings. Latest Advances in the Fast Changing World of Multimedia (Cat. No.00TH8532).

[10]  Yuan Yan Tang,et al.  Transformation-Ring-Projection (Trp) Algorithm and its VLSI Implementation , 1991, Int. J. Pattern Recognit. Artif. Intell..