The Development Of The Textile Fabric Defect Detection System Based On Machine Vision Technology

During the quality control and inspection stage of production in textile factory,the textile fabric defect detection is very important because it has a direct influence on both production quality and economic benefit of the factory. Compared with traditional manual inspection, the textile fabric defect detection system based on machine vision technology has advantages such as low false detection rate and high efficiency. In consequence, this kind of system is more and more popular in both research field and marketplace.This thesis is based on the analysis of function and performance that a textile fabric defect detection system needs, and constructs a hardware platform with CMOS cameras and Industrial Personal Computer. It also develops a human-computer interaction system using VC++ to control the functions of the system. Gabor filter image processing algorithm has been researched and used to accomplish the core algorithm of fabric defect detection. This system uses a touch screen to display information and operation,and controls PLC and sound-light alarm when the defect has been detected. Multi-cameras are used in order to cater various ways of producing textile,and the covers of the width is adjustable.Through the performance tests of the platform and program, some results are obtained. By actual test, the system works under stable and precise condition and conforms the expected performance target. The detection rate of textile fabric defect is above 95% which has met the expectation performance index. Finally, this thesis summarizes the system design and show the future development of the textile fabric defect detection system.