Image Retrieval Based on Quad Chain Code and Standard Deviation?

In this paper we present a new idea for image retrieval depending on Quad chain code and standard deviation. The proposed system consist of two stages ,the first one is constructing a database of image training by dividing each image into number of blocks with size 8*8 pixel based on Quad chain code and the structure of database for each block of image consisting of one record with two columns (first column contain a set of chain code and the second column contain standard deviation of block pixels ).In the second stage the test image is divided into a number of block based on quad chain code and find standard deviation for each block and apply a matching operation for each image in database to find what is the most matching image.