冷轧带钢表面自动监测系统的研究

介绍一种新的基于图像处理的冷轧带钢表面自动监测系统。该系统采用面阵ＣＣＤ摄像头来采集带钢表面的图像，并且配备两种不同的照明方式，以检测不同类型的缺陷。系统在硬件结构和软件流程上进行了特殊的设计，以保证能够对带钢表面进行实时在线监测。同时，讨论了系统中采用的图像处理和模式识别方面的一些有效算法。用实际的样本对该系统进行试验，结果表明：该系统能识别六种常见的表面缺陷，识别率接近９０％。

RESEARCH ON AUTOMATIC SURFACE INSPECTION SYSTEM FOR COLD ROLLED STRIPS

A new automatic surface inspection system for cold rolled strips based on image processing is introduced CCD matrix cameras are used in the system to capture images of strip surface, while two different illumination structures are equipped to detect different types of defects The system is peculiar in design of hardware and software, in order to assure on line inspection of strip surface Some efficient algorithms of image processing and pattern recognition for the system are presented The system has been tested with samples from a cold rolling plant. The result shows that the system can detect and classify six common defects, and the recognition ratio is close to 90 %