Research On The Diameter Measurement System Based On Maching Vision

With the rapid development of computer vision and photoelectricity, the machine vision inspection technology has been rapid development and wide application, machine vision inspection as a non-contact detection metheds, with a high degree of automation, flexibility, and good real-time, etc., has increasingly attracted attention.In this paper, the system architecture of diameter measuring based on machine vision, and then describe the software and hardware composition and their role, and the diameter measurement system’s pre-processing of image, the edge detecting of iamge and other related processing of image algorithms and technology are studied and discussed。A1the same time Hough transform for detecting straight line and circle is often encountered in the diameter measurement and it is studied depthly in this paper。 Because existing Hough transform for detecting straight line and circle is not precise enough, so it is improved in the paper, that is, post-test in the Hough transform, and then carry out least-squares fitted line, so that the improved line detection is not only a certain degree of noise immunity, and having a higher detection accuracy。 A1the same time based on the classified Random Hough Circle detection, it is improved from set of points from the edge to remove noise points, the edge set of points determine a straight line, using the edge point gradient direcion information to reduce the invalid in terms of accumulation of Random Hough Circle, so that detecion more efficiently.In addition this article presents a new calibration method for vision measurement system—calibration method using concentric circles planar template。 The method by adjusting the position and orientation of planar template to the best in order to eliminate the impact of such bias distortion, using the imaging template of concentric circles at the best position to calculation their Average Pixel Equivalent。As the plane of the concentric circles template non-linear relationship between the radius and therefore to eliminate the random errors in the process of calibration, and then using the multiple Average Pixel Equivalent at random locations within the depth of field, then complete the entire visual scence of deep space calibration finally, improve the efficienty of the calibration effectinvely, the new calibration metheod is simple, high precision, and then applied the new calibration method to the measurement of the diameter of cable line, and achieve a higher precision, which is suitable to conduct on-field industrial dimension measurement calibration, has a broad meaning of popularization and application。Combination of the above research results, this paper using VC++6.0software combines OpenCV image library and using the modular idea to developed a machine vision system for measumenting the cable diamerer, and regular cable diameter measument experiment, the experimental results of this system is really satisfied fully able to meet the required accuracy to effectively and effectively the results of the results of research papers and finally the presence of experimental error is analyzed, and proposes some improvements direction。...