Journal of Textile Research ›› 2016, Vol. 37 ›› Issue (05): 150-154.
Previous Articles Next Articles
Received:
Revised:
Online:
Published:
Abstract:
To reduce the errors of fabric color caused by light and experience when measured by human eyes, a color recognition and simulation methods for automatically matching fabric color with special standard color card was put forward. This method is developed based on computer vision and image analysis technique, and fabric color automatic registration system for 1 925 Pantone TCX color swatches was realized. The method used the scanner to acquire Pantone TCX color scan images, and then extracted effective color characteristic information from color image for constructing database of color image feature.Color layered model and registration model of chromatic value and Pantone TCX color card based on "one to one" support vector machine (SVM) were designed. After optimization and implement training of color identifying for SVM model parameters, the accuracy is 96.89%. Using 296 previously unknown samples for verification, accuracy is 98.85%. This would provide a reference tool for objective, fast color measurement and color of digital simulation.
Key words: fabric color, image processing, support vector machine, registration, Pantone color swatches
0 / / Recommend
Add to citation manager EndNote|Reference Manager|ProCite|BibTeX|RefWorks
URL: http://www.fzxb.org.cn/EN/
http://www.fzxb.org.cn/EN/Y2016/V37/I05/150
Cited