边裂是严重影响冷轧带钢质量的缺陷之一,是引起断带的主要原因。介绍了在基于机器视觉的检测系统中用图像处理技术检测带钢边裂的方法。首先通过边界追踪得到带钢边界的曲线链码,然后应用距离累积曲率算法估计链码重构曲线的曲率分布,以曲率的大小和方向为判断条件检测边裂缺陷,进而计算边裂的位置及开口深度。实验结果表明,该方法应用到实际检测系统中能够可靠地检出带钢边裂。 Edge cracking is one of the defects that seriously affect the quality of cold-rolled strip, which is the main reason for the break. The paper introduces the method of detecting the edge of steel strip by using the image processing technology in the machine vision based inspection system. Firstly, the curve chain code of the steel strip boundary is tracked by the boundary. Then, the curvature distribution of chain code reconstruction curve is estimated by the distance cumulative curvature algorithm. The edge crack is detected by the size and direction of the curvature, Opening depth. The experimental results show that this method can be applied to the actual detection system to detect the strip edge reliably.