Clustering and retrieval using self-organized mapping networks based on image color and texture

This paper represents a method to clustering and retrieval images by Self-organizing map which using image's color and local texture information.Color information calculated by spatial color accumulates in the HSI color space,which avoid the high dimension problem.Additionally,using local texture descriptor decrease the inaccurate level which caused by only using color information.The ability of Self-organizing map such as characteristic selection and unsupervised learning are sufficiently used.