论文部分内容阅读
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 %。
A new automatic surface inspection system for cold rolled strip based on image processing is introduced. The system uses an area CCD camera to capture the image of the strip surface and is equipped with two different lighting methods to detect different types of defects. The system is specifically designed in terms of hardware architecture and software flow to ensure real-time on-line monitoring of the strip surface. At the same time, some effective algorithms for image processing and pattern recognition in the system are discussed. The actual sample was used to test the system. The results show that the system can identify six kinds of common surface defects with a recognition rate of nearly 90%.