Petite Feature Set Defining Solid Structural and Color Assets in CBIR Processes

Content-based image analyses and retrieval systems, most of the time, deal with histogram refinement techniques. The coarse characterization of histograms is researched and a very brief list of features is proposed in this paper, which is recommended for RGB images carrying some structure or colored information with similar appearance. The R, G, and B contents are split and their color and edge features are considered. These features include the count of horizontal, vertical and slanting edges, the number of clusters, and the percentage of noticeable colored channel by area. Iteration is committed onto the limited set of results generated by this procedure, which further refines the results based on the overall appearance of these selective images. This reduces the order of complexity, and simultaneously, enhances searching ability by reducing number of computations and comparison time.