Abstract:

In order to discuss an smart evaluation method for objective evaluation on fabric appearance quality, patches extracted from woven fabric images with different densities were used as training samples and discrete cosine dictionary was used as the initial dictionary of learning algorithm based on the least square method. The original woven fabric image samples can be restructured well by the dictionary by a linear summation of its elements. To evaluate the reconstruction performance, mean square error was selected as evaluation index. The influence of gray distribution of fabric images on the reconstruction error was discussed, and then the reconstruction of density on the reconstruction error were discussed with the normalized image gray value. The experimental results show that when the number of dictionary atoms equal to 9, the mean square error of plain increases firstly and then remains within a certain range and the mean square error of twill increases with the increasing of warp and weft density from 150 to 360 yarns/10 cm.

Key words: dictionary learning, woven fabric, texture representation, density