Abstract: An image segmentation method is provided for the colored fiber blends high resolution image, and the segmentation process is carried based on the color characteristics. The whole process is discussed in the HSV color space: the gray region and the color region are segmented based on its value in the V channel and H channel separately with the Fuzzy C-means clustering method (FCM). The test result proved that, this segmentation method based on the FCM is much efficient to separate the image into different parts with the same texture and color hue with the original image. The algorithm can be carried to the computer aided manufacture for the color separation, which will be more directly and efficient than the traditional analysis method based on the optical model.

Key words: mixed dyed textile, fuzzy C-means cluster, Image segmentation, HSV color space