Abstract:

In order to establish a scientific and effective evaluation standard of colorimetry indexes for colored spun yarns, a large-area macro image acquisition system,  based on digital color measurement theory, was designed and implemented to solve the problem of conventional macro image acquisition on limited vision field, providing large area macro images for extracting and analyzing chromatic features. According to the specific coloring mechanism and process of colored spun yarns, a model combined global color features with local texture features for colorimetry index was established, consequently. Extracted third-order color moment features and color local directional pattern features can not only make full use of the complementarity between global and local features, but also combine diversities from different color spaces. Experimental results indicate that the extracted diversity features can accurately and steadily characterize the tiny changes of colorimetry indexes, and demonstrate ideal effectiveness and robustness of the fusion algorithm. Meanwhile, compared with conventional spectrophotometry, the designed novel systemis more adaptive and practical.

Key words: colored spun yarn, colorimetry index, large area macro image, color moment, color local directional pattern, feature fusion