2D shape classification and retrieval

Conference item

We present a novel correspondence-based technique for efficient shape classification and retrieval. Shape boundaries are described by a set of (ad hoc) equally spaced points – avoiding the need to extract “landmark points”. By formulating the correspondence problem in terms of a simple generative model, we are able to efficiently compute matches that incorporate scale, translation, rotation and re- flection invariance. A hierarchical scheme with likelihood cut-off provides additional speed-up. In contrast to many shape descriptors, the concept of a mean (prototype) shape follows naturally in this setting. This enables model based classification, greatly reducing the cost of the testing phase. Equal spacing of points can be defined in terms of either perimeter distance or radial angle. It is shown that combining the two leads to improved classifi- cation/retrieval performance.

Authors: McNeill, G; Vijayakumar, S

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: International Joint Conferences on Artificial Intelligence
• Host title: IJCAI-05: Proceedings of International Joint Conference on Artificial Intelligence
• Journal: IJCAI-05: Proceedings of International Joint Conference on Artificial Intelligence
• Publication date: 2005-08-01
• Acceptance date: 2005-01-01
• Keywords: SBTMR