Color Image Segmentation using ERKFCM

Color image segmentation is an important task for computer vision. The segmented RGB color space is not more reliable and accurate for computer vision applications. For this purpose, the proposed approach combines different color spaces such as RGB, HSV, YIQ and XYZ for image segmentation. The combine segmentation of various color spaces to give more accurate segmentation result compared to segmentation of single color space. The images are segmented using Kmeans clustering and Effective robust kernelized fuzzy c-means(ERKFCM). Two significant criteria namely PSNR (Peak Signal to Noise Ratio) and MSE (Mean square error) are used to evaluate the performance.