Background Removal for Color Images based on Color Components Differences

X. Tian, Y. Zhao, C. Tao, and Z. Tang (PRC)


Background Removal, Image Segmentation, ColorFeatures Enhancement, Color Components Differences


A new algorithm for background removal in color images has been proposed. Instead of using color edges or region color similarity of the same color component to do segmentations in the common approach, the new approach uses color differences between different color components to do segmentation. The components differences features, DGR DBR, have been defined and the new segmentation algorithm based on DGR DBR features has been described. The proposed new algorithm has been implemented and tested. Testing results have showed that the new algorithm is robust, effective and simple as well as bringing very good segmentation results to us.

Important Links:

Go Back