Image Processing Based Color Difference Detection of Digitally Printed Cotton Fabrics
Fabric color difference rating mainly relies on manual comparison of standard sample cards for deter-mination,and its rating is easily affected by human subjective factors.In order to carry out objective and global evaluation of cotton fabric's color difference,it proposed three OpenCV-based fabric color difference evaluation methods.Firstly,we obtained different color difference grayscale images of the standard sample card under D65 light source,used Gaussian filter to reduce the noise of the image,and then quantified the grayscale image of the standard sample card using MSE mean square error,PSNR peak signal-to-noise ratio and SSIM structural similarity index algorithm to establish and find the best model.In order to verify the accuracy,digital printing was used to spray dye cotton fabrics after different pretreatment processes as a test sample,and the color difference level was obtained through the model and compared with the manual evaluation of the color difference results.The results show that the evaluation results of the three kinds of image processing models for evaluating the color difference of fabrics are all in high consistency with the manual evaluation,and the image processing methods are more concise,objective and efficient.
万方数据
维普
