Research And Implementation Of Colorful Printed Fabric Defect Detection

The quality of textile not only is a standard for 20 measuring both production and management level of enterprises, but also affects textile trade. Hence, it is significant that detecting defects while putting fabric into use. With the development of modernization, market value of colorful printed fabric is improving. It is more important to build an automatic detection system based on machine vision to replace traditional manual fabric defect detection system. It is well worth studying and realizing the defect detection. This subject is focus on colorful printed fabric defect detection. The work is indicated as follows.(1) A detection algorithm based on CIE L*a*b* color space is proposed to detect various colorful fabric defects more accurately and avoid the error detection resulted from losing image information of graying. CIE L*a*b* color space is transformed from XYZ space via nonlinear transformation of RGB space. First, three features including the brightness characteristics L*, color information of a* and b* which are used to form 2-D Gabor filter in a complex number can be acquired. Genetic algorithm with the objective function in a form of energy is adopted to select optimal Gabor filter parameters. Second, filter built by flawless image is used to filter the corresponding drawbacks of image. Third, ultimate results are obtained through sliding threshold segmentation method and properly de-noising. The experiment indicated that this method can detect diversiform colorful fabric defect accurately.(2) In order to detect defect of the complex background texture fabric fast and accurately, and decrease computation of traditional algorithm, an algorithm based on sparse coding dictionary learning is proposed in this paper. Firstly, Radon skewness correction is adopted to reduce the pixel information processing error, and then Gabor filter is used to eliminate noise after correction. Secondly, sliding window with certain size is applied to select the image block in preprocessed-image and build input sample matrix. Dictionary and sparse coefficient of flawless sample matrix are obtained via K-SVD. Horizontal and vertical projection feature matrixes are calculated after sparse reconstruction. Thirdly, test sample achieved sparse reconstruction through the dictionary and coefficient that have been treated with flawless image. Projection matrixes of test image is obtained as the same way and used to build corresponding horizontal and vertical feature matrixes to flawless image, respectively. Then spot the defective region of image. Experiments showed that the proposed algorithm could efficiently detect defects with shorter time.(3) The hardware platform and software are introduced in detail among this thesis. Each device is chosen according to parameters that detection system need. Experiments show that detection system proposed in this research is stable, efficient and real-time in industry.