Based On Color Features Of Color Image Retrieval Method

After the introduction of the color-based image retrieval technology characteristics of the preparatory knowledge and key technologies, the papers focus on the color histogram and a gray-scale images used to retrieve the gray histogram algorithm. The traditional colors used in color histogram space do not meet the visual characteristics of the human eye,and only records the image of a loss of the location of the problem of information, made the following improvement: Use of visual consistency of the HSV color space instead of RGB color space; The different colors for different channels to quantify the level of H,S,V three components,such an spacing of non-quantifiable; To H,S,V weighted sum of three components that the largest area of color histogram of color. Reuse images in different colors formed by the connectivity of the region's largest area to replace the color of the spatial distribution of the total number,constructed the largest area of color histogram.Finally using Visual C++ to improve the algorithm was verified by check-rate and the recall rate for evaluation indicators to prove that this method is effective.