The Strip Steel Surface Defects Detection Adaptive Preprocessing Method Based on Double Field

Abstract

Double field detection method can effectively solve missing detection and mistake detection problems. But because the image data is two times as single field detection, and therefore the detection speed is slow. A new Processing flow is put forward. Firstly, the suspected defect images are filtered out, and only the images that contain defects can be processed in the follow programs. Secondly, the retained suspected defect images are divided into two categories. Different categories are handled by different process. The simulation experiment result shows that: it not only makes full use of the advantages of the double field method, and greatly reduces the image data quantity, at the same time ensuring the real-time of the system.