The Detection Of Digital Printing Fly Line Based On The Hough Transform

With the rapid development of Chinese society, it put forward higher requirements on the performance of the market quality and equipment of textile products, automatic defect detection system is gradually replacing artificial detection of digital print fabric. China is the country that produces a lot of textile production, but the textile productions are not very competitive. Although we can buy defect detection system from abroad, the price is too expensive, the configuration and operation mode of the machine can not adapt to a series of problems in Chinese specific production situation. The progress of production technology lead to the production efficiency of make a spurt of progress, but the quality of detected technology of digital printing fabric has lagged behind the technology of the production, become a bottleneck affecting labor efficiency in the textile industry. Naked-eye detection is not only the high false positive rate, low efficiency, but also long-term repetitive work can cause workers tired, automatic detection and has been put on the agenda. With the development of computer technology and image processing technology, it may be possibilities that the computer vision digital printing fabric defect detection and based on the recognition. It not only can enhance the production efficiency of textile that using the computer vision of digital print fabric defects identification, but also develop objective standards of the type, size and product quality assessment for digital printing fabric defects to lay the foundation.The so-called fly line, refers to some horizontal stripes in the digital printing process, is one common defect in digital printing products, whether that can be tested or not affect the quality of the products, and therefore it become the focus of this paper.At first, this paper first introduces the status of research and application of digital printing fabric defect detection, and the system of fabric defect detection system consists of the hardware and software environment, and then put forward by the Hough transform based detection method, correlation coefficient for the theory of digital printing fabric, provides the theory reference defect detection system for digital printing products the next step in the development of suitable for China’s reality.Aiming at detect flying line for digital printing in digital printing process, templates can be used between the fabric and the method of calculating the correlation coefficient test, but the efficiency is not high. As our products are mainly detected in common with linear characteristic defects, it can improve efficiency that the feature of the line based on Hough transform and then the test of the correlation coefficient.This paper presents a technology based on Hough transform to build a automatic detection system of flying line for digital printing, combined with the image processing technology, using MATLAB7.10for digital printing digital printing fabric defect detection model, a linear feature information of image in the extraction by Hough transform, and then detected flying line the digital printing process by calculating the cross-correlation coefficient.Finally the technology can fairly well detect flying line for digital printing through comparing with manual inspection.