Abstract:

In view to the problem of accurate detection of printed patterns of textiles based on the machine vision, this paper presented a method of applying the JSEG algorithm to the image segmentation of the textile printing pattern image, and completion of image region segmentation included image filtering, color vector, regional growth, and merger. The impact of the threshold value of color quantization and the threshold value of region merging on image segmentation and method selection was investigated by experiments. Finally, the results of image segmentation of textile printing pattern by using this algorithm were demonstrated, indicating that the algorithm which uses color information and regional information is efficient in image segmentation of weave information contained textile printing pattern, and shows certain inhibiting effect to some interference caused by the sun light.