Research On The Methods Of Image Content Retrieval Based On Shape Contexts

As the development of multimedia information technology and Internet technology, especially after the appearance of mass image database, more and more attention have been paid to image information. How to find a special image from the mass database efficiently and effectively has become a major issue to be solved. In this case, the research on content-based image retrieval was developed, and became a hot spot in the fields of multimedia information processing. The shape retrieval based on shape is of the best research value and practical significance in the image retrieval, meanwhile, there are many problems to be solved.The shape feature extraction and similarity measure are the core issues of the shape retrieval based on shape, these two issues have been widely studied. The shape contexts, as algorithm of shape retrieval based on shape proposed in recent years, with a unique shape descriptors and similarity metrics, achieves good experimental results in image matching and retrieval, by a large number of experts and scholars.Through deep understanding and study of the shape contexts, we propose a dynamic selection of feature points based on shape complexity. This can take full advantage of the shape feature to be retrieved, as to dynamically select , rather than a particular value of the fixed set. To the relatively simple shape of the images, it can effectively reduce the search time. Taking into account the different characteristic points playing different role on the overall shape, we propose the energy function of curvature, It is assigned to give each point the corresponding energy weights, Experiments show that this retrieval algorithm has greatly improved performance. Finally, for the shortcomings of the missing profile, we propose a shape matching based on the shape coefficient histogram. It can overcome the impact of missing profile and obtain accurate matches.