Application of a New Image Recognition Technology in Fabric Defect Detection

At present, defect detection during the manufacturing are still finished by man, there are a lot of weaknesses by this way, such as low detection efficiency, high miss rate. All of these affect the production quality seriously and restrict the further improvement of production efficiency. This paper presents a method using Fisher classifier in computer image pattern recognition for defect detection and grade scoring of fabric, and gives the realization of software programming and testing. The test results show that the defect recognition rate is 94%.