Based On Neural Network For Rapid Detection Of Steel Surface Defects

The surface defect is an important factor influencing the quality of steel and the current research of this field focuses on the rapid defect inspection in the steel surface at home and abroad. A study into the technology of surface defect recognition is of great significance both in theory and practice. Inspection and classification system of defects in the surface of cold rolled strip steel is designed. The research focuses on the application of BP Neural Network Method and image processing technique to the recognition of steel surface defects, which realizes the rapid auto-classification of defects in the surface of coldrolled strip steel. The results of the research are as follows:1. According to the strip surface detection system needs to adopt a relatively reasonable structure of the detection system, the system by the CCD camera 8 and 8 parallel processing computer. The results showed that the stability of the system and maintenance of a strong, and adaptable.2. Based on the practical situations in which defects occur in factories, the thesrs analyses the programming requirements of the steel surface inspection system and presents new software designed to serve as a steel surface inspection system.3. The research develops the application of neural network to the defect classification of cold rolled strip steel. Based on the different characteristics extraction, a method of steel surface inspection based on BP Neural Network is proposed by adopting multi-classifier technology. This design realizes a steel surface defect recognizer based on neural network. The experimental results show that the defect classification method can recognize the types of defects in the surface of cold rolled strip steel and the recognition rate is up to 96%.The paper studies the application of the image processing technology, the neural network and model recognition theory to the surface inspection, and successfully realizes the auto-inspection and recognition of the defects in strip steel surface. It can meet the requirements of surface inspection for the strip steel assembly line, and is of great practical values.