We present a new method of retrieving images from an image database. We combine both color and spatial features of a picture to index and measure the similarity of images. Several color spaces that are widely used in computer graphics are discussed and compared for color clustering. In addition, we propose a new automatic indexing scheme of image databases according to our clustering method, which could filter the image efficiently. As a technical contribution, a Seed-Filling like algorithm that could extract the shape and spatial relationship feature of the image is proposed. Also, the system is incorporated with a visual interface, which allows the user to specify objects as the spatial specification of pictures. Our experience shows that the system is able to retrieve image information efficiently by the proposed color-spatial approach
Proceedings of The 2nd International Workshop on Multimedia Network Systems (MMNS'2000), Iwate Prefectural University, Iwate, Japan, July 4 - 7, 2000.