针对圆筒形零件直角梯形槽槽宽的工业检测速度和精度要求,采用工业相机、镜头、计算机及专业图像处理软件等,构建了一套机器视觉系统,并依据图像处理算法设计了一个针对圆筒形零件上直角梯形槽槽宽检测的方法。 对比实验表明,该系统的检测精度能够达到微米级要求,明显优于人工采用专用槽宽测量工具的测量精度,且测量结果不受主观因素的影响,效率较高。

A machine vision system was constructed using industrial cameras, lenses, computers, and specialized image processing software according to industrial inspection speed and accuracy requirements of groove width detection of right-angle trapezoid groove of cylindrical part. Groove width detection method of right-angle trapezoidal groove of cylindrical parts was design based on image processing algorithms. Comparative experiments showed that the detection accuracy of the system in groove width of cylindrical part can achieve micron order, which is better than manual measurement using a dedicated groove width measurement tools; the measurement results are not influenced by subjective factors; the efficiency is very high.