Shape-based image retrieval using spatio-temporal relation computations

Content-based retrieval of multimedia information is one of the most difficult research topics in multimedia computing and information retrieval. In this paper, we present a visual system which allows content-based retrieval of still image. The recognition algorithms we used are based on spatio-temporal relations. Two approaches and algorithms were developed based on the similarity between polygons. The system is incorporated with a visual interface which allows the user to specify polygons as the shape specification of pictures. The preliminary experience shows that, within an image database containing about 300 bitmapped images, the system is able to retrieve correct information of a high satisfaction.