A Device And The Technology Of Color-difference Defection For Denim Jeans In The Washing Process

In recent years, with the rapid development of the denim business in China, the annualexports of denim clothing is the first in the world. In spite of China is the largest producer ofdenim business, it is far from the powerful nation of denim business. Because there are asmall number of high-end quality products. Through the analysis of the main reason, the mostimportant reason is that various standards of the textile products testing are imperfect. Thedetection method is not scientific enough resulting in some companies product many inferiorproducts for profit-making. This article studies device and technology of denim washingprocess color-difference testing, it use computer vision instead of human eyes and it is morefast and convenient. This system will realize the aim of the washing denim clothing’scolor-difference testing which is objective and rapid. Also this system will play a guiding rolein the adjustment of the washing process.The main contents include five parts. In the first part, it introduce the researchbackground, significance, color-difference detection and related fields from the domestic andinternational. Also it recommends the current situation and the trend of the denim clothingcolor-difference detection field.In the second part, it expounds the overall design the hardwareof the system. In the third part, several kinds of color space models are summarized. Andgraphic denoising theoretical approaches are introduced. It used MATLAB and VC mixedprogramming techniques which achieved the original image preprocessing in MATLAB andconverted to VC project file. In the fourth part, it recommends the color space conversionmodel in detail, and some common color difference formulas are introduced. Based on thecomparative research on the color difference of different color spaces, I found out somepractical color space and color difference formulas. I list the reasons for selecting CIEDE2000color difference formula. In the fifth part, I summarizes the advantages and disadvantages ofthe color rating indicators and methods, and propose a similarity index to quantify color rating.The similarity--color-difference value curve which is more in line with human visual wasfitted. With the help of the MATLAB Statistics Toolbox software, the feasibility and accuracyof the color-difference testing system was proved. In the last part, I summarize the results ofthis work, and point out the development direction for future research. It is of great significance for the fast and accurate detection of denim which computermachine vision applied to the denim color-difference detection. There are two aspects ofusing this system. Firstly, instead of subjective evaluation, objective evaluation could reducethe generation of error and reduce the strength of the testing staff. Secondly, it could improvethe testing efficiency greatly. And in order to achieve the goal of improving the accuracyabout the color-difference testing, the testing samples could be increased.