Dimensional inspecting system of shaft parts based on machine vision

In this paper, a vision based measurement system for shaft parts is presented. The hardware of the proposed system consists of image acquisition system, lighting system, industrial control computer and measuring work frame. The key of the whole software system is the image processing algorithm, and its performance directly affects the accuracy and robustness of the measurement system. Firstly, the image pre-processing is used here to eliminate noise, and then the region of interest of the image and the spindle direction of the work piece is determined by using rotating projection method in order to reduce the computational cost of image processing and improve the realtime performance. Then the maximum and minimum gradient is computed to get the left and right edge of the work piece. By using bilinear interpolation, the sub-pixel edge locating accuracy is 0.2 pixel to improve the measurement accuracy. Then, Hough transformation is adopted here to calculate the distance between two edge lines. Finally, the calibration experiment and the measurement experiment of the work piece are carried out using 1.3 million pixels industrial camera, and the measurement error is analyzed. Experimental results show that the measurement accuracy can reach to 0.015mm, so the system has high measurement precision. At same time, it is very robust to rotation, scale change of the work piece and noise, so it can meet the requirements of the measurement of shaft parts.