| This thesis is based on the Chinese Textile Industry Association project “dyeingproduction process of online automated testing system”. In view of China’s dyeingand printing industry present situation, the research of continuous hot melt dyeingmachine dye aberration on-line detection system, to solve the continuous melt dyeingmachine of dyeing fabric Color-difference detection problem in China, greatlyimprove the quality and efficiency of printing and dyeing fabric.Continuous pad dyeing with high production efficiency, adapt to the wide variety,process control is simple and many of the printing and dyeing factories employed, butits left,midst and right color-difference is the long-term problems,even beyond repair,affect the quality of the products. The main topic on the left and right color-difference,the working principle of the system: above dyeing cloth is mounted the threeindustrial cameras, to measurement the left,midst and right color differencerespectively. Camera under certain lighting conditions to capture an image; the imageacquisition card image signal digital storage into the industrial computer;computeraccess to shoot the color information of the image, with the set of samples of colorinformation is compared to the calculation of color-difference. Once thecolor-difference value exceeds the required range, namely warning monitoringpersonnel real-time adjustment, prevent the cloth to present the color-difference.This paper based on the PC machine vision system to study. In recent years,machine vision in industrial applications and achieved great success, the applicationof machine vision system not only greatly improves the detecting speed and range,and can be the product in the certain degree to make consistent evaluation, in order toimprove the detection precision and reliability. In printing and dyeing production line,eye for color measurement and judgment due to fatigue, differences betweenindividuals and cause errors and mistakes, but the machine will tirelessly, stable.Machine vision system includes a lighting system, a lens, the cameras and the imageprocessing system. The fabric online detection of color-difference, need to consider the speed of operation of the system and the image processing speed, the use of coloror monochrome camera, target detection, the position of field, need more highresolution and contrast require etc.. Machine vision is a relatively complex system,the majority of control object is a moving object;the system and the moving objectmatching and coordinating action is particularly important for each part of the system,so the action time and processing speed has strict requirements. The light source, alens, a CCD camera and image acquisition card as part of the visual system imaging,the imaging quality directly affects the accuracy of color-difference detection, thispaper focuses on this study.Color-difference detection is a necessary condition for extraction of fabric colorfeature, and application of color difference formula. Color theory is the base ofcolor-difference. This paper studies the widely used in industry for several colordifference formula;compares the various color space. On the basis of this study hasbeen designed based on the different color-difference formula algorithm, which iscompleted based on different color space.The results are obtained. In severalalgorithms, the two properties are discussed. Through a variety of different materialsand different colors of the fabric color-difference detection, and draw the conclusionthat the most practical textile industry of an algorithm,work to the fabric imagepreprocessing in the RGB color space and extract color feature value, then convertedto CIELAB color space by the color-difference formula CMC (2:1) computation thecolor-difference of fabtic. A large number of experiments show that this method caneffectively detect the fabric color-difference. |
PDF Full Text Request