A novel method for shape representation

The expansion of multimedia information has generated requirements for easy and effective access to large image databases. Content Based Image Retrieval (CBIR) has become a major area of research and has received increasing attention. Shape representation and description is one of the major problems in (CBIR). There are generally two types of shape representations: Regionbased and Contour-based techniques employed in the field of shape retrieval. This paper provides a summary of these two Shape representation techniques. We will propose our novel Structural–based shape representation method which falls under contour-based techniques. It is worth mentioning that the proposed method is a part of our project which is image retrieval based on shape using the user’s sketch as a query input.

[1]  Timothy K. Shih,et al.  Content-based image retrieval with intensive signature via affine invariant transformation , 2000, Proceedings International Symposium on Multimedia Software Engineering.

[2]  Sungyoung Kim,et al.  Usefulness of boundary sequences in computing shape features for arbitrary shaped regions , 2002, Object recognition supported by user interaction for service robots.