Cold-rolled Strip Surface Defect Detection System Design And Classification

Defects on the surface of cold steel strips are main factors to evaluate the quality of cold strips, the surface defects inspection has been studied extensively in the recent past. The technology of surface defects inspection has great application prospect. According to the steel company's requirement about the strip steel's surface quality, a system of automatic surface defects inspecting was developed. In this paper, the algorithms for defect classification based on neural network and image processing technology are discussed. The achievements are as follows:1. According to the requirements of the inspection system and the character of the inspection system which is used currently, an advanced inspection system is Proposed. After the improvement of the system, the number of camera was reduced from six to one, and the number of computer can be changed flexible. The result shows, the system stability and compatibility was enhanced, and the cost falls remarkably.2. The utilizing of parallel computing technology in the cold-strip steel surface defects inspection system was studied. the result shows that the stability and the flexibility of system are improved.3. According to the situation of defects which was provided by the cold-rolling mill, the demanding of the defect inspection system was analyzed and the software was designed.4. Algorithms for defect classification based on neural network are developed. A new method of cold-strip steel surface defect recognizing is proposed based on the different characteristic extractive technology and the multi- classifier combine method. The result shows that the defects inspection correctness enhanced 1%.Experiments with samples of surface detects show that classification rate is up to 97%.