机器视觉检测滚子表面缺陷

Applying machine vision technique to inspect surface defects of rollers

介绍利用机器视觉检测滚子表面缺陷的实验装置。该装置采用512位SSPD列阵作为摄景器件。藉助专门机械装置实现滚子表面全扫描展开,把获得的二维图象信息送PC/AT微机进行图象处理。文中提出了三种图象处理算法:多段平均法、二次背景处理法和局部算子处理法。用上述算法较好地解决了不同精度滚子的表面缺陷检测。附图16幅,参考文献5篇。

A experimental device for surface defect inspection of rollers by machine vision technique is introduced in this paper. In the device, a 512 bits SSPD array is adopted as the panoramic shooting apparatus. The surface of roller can be fully scanning expansed by special mechanism device and the two dimension image information gained will be sent to PC/AT microprocessor for freatment. Three types of image treatment meathods are introduced as follows; multi-segment average method, secondary back ground treatment method and partial operator treatment method. Using above calculation methods, the inspection of surface defects of rollers with different accuracy can be better solved.