基于机器视觉的弧形件检测算法

为快速准确地对弧形零件特征参数进行检测，提出了一种基于机器视觉技术检测弧形件的各种特征参数的方法。该方法先采用Canny算子检测图像边缘，然后用三次样条插值方法对图像边缘进行插值并计算其亚像素坐标，最后采用基于切线方向的方法计算曲率值。对离散曲率数据进行处理，计算其均值和方差以及弧形件的弧长、夹角、面积等特征量。实验结果证明：该算法不仅精度高、速度快，而且稳定性好，能满足弧形零件特征参数的检测要求。

Inspection algorithm of arc-parts based on machine vision

In order to achieve fast and accurate inspection on the feature parameters of the arc shaped piece, the paper presented a method that inspected feature parameters of the arc-shaped piece based on machine vision. First, the image edge was extracted by Canny operator and sub pixel coordinate was computed by a cubic spline interpolation algorithm. Then, discrete curvature was computed by discrete curvature computing method based on tangent direction. The discrete curvature data were arranged; the mean and variance, as well as the arc-length, surface and angle of curved pieces were calculated. The experimental results indicate that the proposed method is of not only high precision, high-speed but also good-stability, and can satisfy inspection requirement of arc-parts feature parameters.