Research on the Precision Measuring System of Steel Bearing Aperture Based on Machine Vision

摘要

A new measurement system was designed in order to improve the accuracy of the non-contact type high-precision measurement of Steel bearing aperture. Firstly, the median filter preprocessing algorithm was designed to remove the impulse and salt-pepper noise in the image, and the edge details of the image was also preserved. Secondly, the image processing equalization algorithm was applied to improve the image information contrast, which the gray level of original image has been extended to the entire gray level range of the output image. Thirdly, the improved canny edge detection algorithm was applied to detect edge more accurately, which employed the convolution operation of the smoothing filter impulse response of first order differential and the original image. The noise suppression and robustness of the operator were improved. Finally, the contour image of the measured part was filtrated to calculate its size by the aperture image detection algorithm. Its system algorithm software was designed based on the machine vision and computer control technology, which used the Visual C++ development language in the integrated development environment of OpenCV. The results show that the precision measuring system is highly accurate and efficient for the parts size measurements. It is helpful to improve the accuracy of the Precision Manufacturing Technology and Measurements.