Study On Color Difference Detection Of Textile Fabric Based On Machine Vision

With the improvement of consumers material living standards,their consumption philosophy has also changed,especially the preference for more comfortable,more beautiful appearance of clothing.The diversification of consumer demand for clothing makes the quality of textiles more demanding,textile enterprises in order to adapt to the development of consumer demand,and constantly introduce new technologies and new methods to improve the quality of textiles.In textile enterprises,the detection of textile fabric color is the main factor affecting the quality of textiles,accurate color feature information is particularly important for textile enterprises.Some textile enterprises still use the human eye to judge the degree of chromatic aberration of printing and dyeing textile cloth,has been far from the merchant requirements,which requires more stable,repetitive and accurate color detection methods to replace the human eye detection.Through the continuous efforts and exploration of scientific researchers,the rapid transformation from the subjective human eye assessment to the objectivity of the color measurement instrument detection,the use of machine vision instead of the human eye for measurement and judgment,can achieve higher detection accuracy and efficiency,for the textile enterprises to produce flexibility and automation development laid the foundation,and become an increasingly effective method.However,because the machine vision detection technology is not advanced enough,using CCD industrial camera with optical lens to obtain the image information of textile cloth,some external unstable factors produce noise points in the obtained image,which affects the original color characteristic information.In order to extract more accurate image color feature information and strengthen the application of machine vision in the detection of color difference in textile cloth,this paper designs the machine vision system,puts forward the color feature extraction method,obtains the accurate color characteristic information successfully,finally carries on the comparison experiment analysis to the chromatic aberration result,can improve the accuracy of textile cloth color difference detection,for the enterprise in the textile cloth testing has a certain reference significance.The main research work and results of this paper are as follows:(1)Textile cloth color differential detection machine vision system design.In this paper,the composition and working principle of machine vision system are studied,the influence of acquisition system and lighting system on obtaining image color characteristics is considered,the selection of camera and lens is completed,the light source and illumination scheme are determined,and the machine Vision acquisition platform is designed.(2)Color extraction of textile cloth based on machine vision.Aiming at improving the accuracy of color feature information,a color feature information extraction method is proposed,including image acquisition,image segmentation,image preprocessing and color feature extraction.A set threshold image segmentation method is used to complete the image acquisition.Aiming at the distortion phenomena such as color deviation,noise point and color information point discontinuity,the image is preprocessed and the color information is adjusted.In this paper,an improved automatic iterative adjustment Gamma color bias correction algorithm is proposed to obtain the true reduction of Image color information,and a color histogram equalization based on HSI color space is used to improve the overall contrast of image color.An improved filtering processing algorithm is proposed to remove the texture features of redundant noise points and textile cloth,and finally the color feature extraction of image information is accomplished by using an efficient K-means clustering algorithm.The results show that this method can effectively remove the influence of irrelevant information points on color feature information and extract more accurate color feature information.(3)Color characteristics and chromatic aberration evaluation and analysis.The computer detects the absolute color,while the human eye detects the relative color.Therefore,the use of the computer in color evaluation,the color of the detection needs to be adjusted according to the visual effects of the human eye,to ensure that the color characteristics of the detection and human eye vision to maintain a good relationship.Through the analysis of color characteristics and the understanding of the theory of color difference of textile cloth,the color characteristics of CIELAB color space are selected,which is close to the visual characteristics of human eyes.The extracted color feature information is brought into the common chromatic aberration evaluation CMC(l:c)chromatic aberration formula,CIE94 chromatic aberration formula,CIEDE2000 chromatic aberration formula and CIELAB chromatic aberration formula to calculate chromatic aberration results.In this paper,the image acquisition of monochrome textile cloth is completed on the designed machine vision system,and then the color feature information of the collected image is extracted by image segmentation,image preprocessing and feature extraction method,and finally the extracted color feature is brought into the chromatic aberration formula for the evaluation and analysis of chromatic aberration results.The results show that this method can better remove the texture features of the textile fabric surface in the collected image and the irrelevant noise points in the image,make the color feature information closer to the standard value,significantly improve the accuracy of the color characteristics of the image,and provide reference for the automatic detection of textile fabric color in textile enterprises.55 maps,6 tables,and 85 references.