An Efficient Statistical Content-Based Color Image Indexing and Retrieval

The world has recently witnessed a tremendous development in the various fields of information technology. Traditional databases do not provide a satisfactory environment for variant media types such as audio video and still images. An efficient statistical image indexing technique has been developed. The mean, variance, and skewness of the image color histogram distributions were used as a feature vector to tag images in the database system. The RGB color space was used. Therefore, the feature vector formed from the mean variance, and skewness of the three basic components of the color space. It has been shown using statistical analysis that the statistical feature vector entries give sufficient knowledge about the histogram distribution for image indexing and retrieval systems. Hence, The proposed system has shown its superiority over traditional histogram based systems in term of feature index properties and length as well as performance.