Fast Color Automatic Measurement Method And Implementation In Textile Color Management

In the textile industry,the quality of color processing has attracted more and more people's attention.The emergence of color management has greatly improved the quality control and communication management efficiency of enterprises.As we all know,high quality color management based on a large number of color measurement data.However,the existing color measurement method can not avoid the influence of textile fabric deformation and the measurement data error is large.In order to meet the increasing requirements of color management in color management in textiles,this paper systematically studies the automatic measurement method of textile color and its key technologies.Using hyperspectral imaging technology,non-contact one-time scanning imaging of textiles is carried out,and accurate and complete color information is extracted from spectral data by computer processing.The way of fast color automatic measurement of textile color management based on hyperspectral imaging technology is achieved.Based on the analysis of textile color measurement requirements and characteristics,this paper studied and determined the selection and construction schemes of components such as spectrometer,CCD camera,lens,light source and transmission mechanism in hyperspectral imaging system.The system was also tested,and the test results showed that it had good spectral measurement accuracy and stability.In order to realize the automatic positioning of the color blocks in the deformed fabric,the distribution structure of the color table was redesigned.At the same time,the methods of hyperspectral image acquisition,image processing,data extraction and related theories of color table were also studied.The preprocessed hyperspectral image locates the center of each patch mainly by using computer template matching and thin plate spline interpolation algorithm.Based on this,the value area of each color block is automatically demarcated and color information is extracted.Research shows that:The method is mainly based on hyperspectral imaging technology,using non-contact one-time scanning imaging of textiles with hyperspectral camera,and extracting accurate and complete color information from spectral data through computer processing.Compared with the traditional color measurement method,this method has high measurement speed,good stability,can effectively eliminate the influence of fabric deformation,avoids various defects of the manual measurement method,and greatly improves the color measurement efficiency while ensuring the accuracy of the color measurement data.For solving the problem of sample staining,an intelligent de-noising algorithm is introduced in the process of extracting color spectrum data,which is used to eliminate the data of small areas such as dirt,stains and wrinkles on the color blocks.And the accuracy of the final color data is ensured.In view of the other measurement error causes the color to jump and invert,which leads the following generated ICC Profile to output problematic color that involves large color error and serious light and dark jump.In this paper,a multi-directional curve fitting algorithm combines with the weight coefficient corresponding to each direction curve is proposed to optimize the bad points automatically with excessive error in the measurement data.By analyzing the application effect of the optimization method through the relevant experiments,in the end,the results show that the multi-directional curve fitting optimization algorithm is automatically optimized for large error data,has little impact on other normal data.The method can also effectively reduce the impact of unforeseen unexpected errors in the measurement data.By comparing the optimization results of 25 sets of traditional measurement data,it was found that:the overall average color difference of the ICC Profile generated by the optimization is reduced by 12.30%;the average color difference of the difficult color is reduced by 16.67%;the average color difference of the neutral gray is reduced by 16.74%.To a certain extent,this method has improved the color transition fluency of textile digital printing.This verifies the effectiveness and versatility of the data optimization method.By comparing the measurement results of the hyperspectral imaging system and the traditional color measuring instrument X-Rite il in this paper,the results show that the stability of the hyperspectral imaging system is better than that of the traditional color measuring instrument X-Rite il under the condition of the same color measuring quality.Therefore,a fast automatic color measurement method based on hyperspectral imaging technology in this paper provides a new way for the field of textile color measurement.