Evaluation Effect Of Color Difference Formulas Based On Polyester Plain Weave Fabric

Color Difference is an important indicator of product quality assessment and classification in the fields of textile,printing,automobile,imaging,and medicine.For the color difference values calculated by different color difference formulas are different,and the effect of using different color difference formulas to evaluate the color difference of textiles is also different,so we need to calculate whether the color difference values calculated by the existing color difference formulas are consistent with the human eye on the textile color.This project builds a small and medium color difference data set based on Sino color book,and a very small color difference data set based on self-dyed fabrics.Four main color difference formulas（CIELAB,CMC,CIEDE2000,CAM16-UCS）are used to carry out a comprehensive and systematic evaluation of the consistency between the calculate and vision color difference of the textiles,and the discernible color threshold and acceptable tolerance corresponding to each color difference formula are also obtained.The main research contents and conclusions are as follows:（1）For medium and small color difference,this thesis uses the Sino color book（SCB）as a small and medium chromatic aberration data set,which has the largest number of color samples made by polyester plain weave fabrics in the world（nearly 20,000）,with good visual equidistance.More than 30 thousands pairs of adjacent color samples in SCB are selected from three directions of Hue angle,Chroma and Depth respectively to analyze the uniformity of four color difference formulas by programming Matlab and statistical analysis of unevenness,PF/3 and STRESS value.The results show that the conclusions obtained by different evaluation indicators in different directions are not the same.Comprehensive evaluation,in terms of uniformity:CAM16-UCS>CIEDE2000_（2:1:1）>CMC_（2:1）>CIELAB,CIEDE2000 is slightly better than CMC;（2）For the small color difference,this thesis builds a small color difference data set（100 color samples）based on dyeing polyester plain weave fabrics with five color centers（gray,red,yellow,green,and blue）.Use the chromatic parameters measured by X-Rite Color i7 to calculate the color difference values of various color difference formulas.After screening,12 observers with a certain color science-related expertise were trained to conduct a Past-Fail subjective evaluation experiment,and then the visual color differenceΔV was calculated using the Z-score method to establish a color difference data set.The WD method was used to statistically analyze the accuracy and repeatability of the observers.The results showed that the average repeatability of 12 observers in the distinguishable color threshold experiment was12%,and the average acceptable tolerance experiment within an acceptable tolerance was 8%.The average value of the accuracy of the color discrimination threshold experiment is 21%,and the average value of the acceptable tolerance experiment is16%.The above results are within the acceptable range.Finally,the degree of consistency between the results of the color difference formula and the results of visual evaluation was calculated by the Stress and F-test methods,and the predictability of each color difference formula was studied in depth.The results show that the consistency,that is,the prediction performance of CAM16-UCS and CIEDE2000 are better than the other two formulas.Among them,the prediction performance of the color difference acceptability of CAM16-UCS,CIEDE2000 and CMC are significantly better than CIELAB;the discernible color threshold and acceptable color difference of the CAM16-UCS are 0.21 and 0.75,respectively,The color threshold and acceptable color difference are smaller than other formulas,which are 0.20 and 0.65,respectively.In summary,the consistency between the calculation results of the four color difference formulas and the visual color difference varies slightly with the evaluation object and evaluation index.From the comprehensive evaluation of the small and medium color difference and very small color difference data sets:CAM16-UCS>CIEDE2000>CMC>CIELAB.CAM16-UCS and CIEDE2000 have a relative certain extent improvement to the CMC which commonly used in the textile industry.In the future,if more and perfect color difference data sets be established based on textile samples,it will be more convenient for the textile industry to optimize or establish better color difference formulas.