论文部分内容阅读
采用图像识别技术对微小塑料齿轮进行质量检测,针对缺齿、齿歪、披峰等齿形误差的随机性,运用三点定圆心法实现塑料齿轮内圆的粗定位,试验数据说明其运行时间比传统Hough变换大大缩短,并为亚像素定位提供了原始数据;提出的降维灰度矩法与最小二乘圆法结合,实现内圆圆心的亚像素定位,同时运用3σ原则迭代法提高了亚像素定位精度;用“虚拟圆扫描法”实现大小轮齿的齿形检测。结果表明该方法可以满足齿轮的自动检测。
Using the image recognition technology to inspect the quality of the tiny plastic gear, aiming at the randomness of tooth shape error such as tooth missing, tooth crooked, and peaking, the inner position of the plastic gear was coarsely positioned by using the three-point centering method. The test data showed that the running time Compared with the traditional Hough transform, the original Hough transform is greatly shortened and the original data is provided for the sub-pixel location. The proposed reduced-size gray moment method is combined with the least-square method to achieve the sub-pixel location of the center of the circle. At the same time, Sub-pixel positioning accuracy; with “virtual circle scanning method” to achieve the size of gear tooth detection. The results show that this method can meet the automatic detection of gears.