Research on Adaptive Main Colors Extraction Algorithm for Natural Images

A framework of adaptive main color extraction algorithm for natural image is proposed in order to make the artistic effects of oil painting on computer much more similar to the hand-printed style in color. As for a natural image provided by a user, the algorithm first transforms the color space of the image into the Lab space to avoid the color distortion caused by the RGB spaceprocessing. Then, the natural image is preprocessed through size compression, image enhancement and denoising in order to eliminate the interference information and preserve the detail information to a better extent in the image.Finally, the silhouette contour factoris effectively combined with the traditional K-means clustering algorithm to extract optimal main colors set in the natural image. A quantity of experiments have been conducted on the natural imagesobtained from a variety of natural environments to test the effects. The results show that the algorithm can effectively extract the main colors in a natural image and the maincolors display the representative color features in the original image in a better way.