A Fuzzy C-means Clustering Based on Hybrid Color Space

The Fuzzy C-means (FCM) clustering is simple and easy to implement in color image segmentation, but firstly, some problems should be solved such as the number of calculating, centers and large computation. So we have designed an improved FCM algorithm based on hybrid color space and firstly transformed the color image from RGB color space into HSV color space, where the number of clustering can be calculated, and after the initial cluster center is calculated in CIELAB color space, FCM algorithm will be used for segmentation. Experiments show that the improved approach can reduce the number of iterations of FCM and greatly enhance the speed of segmentation, which has been applied to the fast segmentation of the colorful wall and floor tiles image successfully.