Research On Detection Technology Of Multi-circle Workpiece Based On Machine Vision

Machine vision detection technology,with the characteristics of non-contact,high efficiency and strong practicability,has gradually replaced the traditional manual detection method,and has been widely used in the manufacturing industry.In the manufacturing industry,circles are ubiquitous in all kinds of workpiece.Circular identification and position detection are of great significance to the production and installation of workpiece.In this paper,the toothbrush pores with multi-circle and small diameter were selected as the research object.The circle recognition and detection are accomplished from the hardware platform of machine vision detection system and the realization of circle detection algorithm.The main contents are as follows:First,according to the characteristics of the research object,the multi-circle detection system based on machine vision is designed.By studying the vision hardware,choosing the appropriate hardware equipment to build the experimental platform,the whole algorithm flow of the detection system is designed.Secondly,the image preprocessing is analyzed and improved.the Laplace operator is used to enhance the gray image to highlight the edge of the workpiece.Aiming at the defects of traditional Canny algorithm in edge detection,In this paper,the median filter which introduces the edge pixel judgment mechanism is used instead of Gauss filter.and only the non-edge pixel and the noise point are filtered.The gradient is solved by using 3 × 3 neighborhood.The gradient amplitude is calculated by adding 45 degree and 135 degree to the original direction,so that the edge can be detected in multiple directions.In order to improve the adaptive ability,the OTSU is used to select the threshold instead of the manual threshold.The improved Canny algorithm can effectively remove small defects and preserve hole edge completely.Thirdly,make a study of circle detection algorithm in machine vision,a circle detection algorithm based on Hough transform and least square approach is proposed.The Hough transform is used to extract the coordinates of the circular edge pixels,andthe least square method is used to fit the coordinates to get the circular position parameters.In order to solve the problem that the Hough transform computation is too large to detect multiple circles,and the edge pixel is sampled by 75%.At the same time,constraining the range of the circle diameter,and invalid cumulative is decreased.Finally,the software of the detection system is developed,and two kinds of multi round workpieces are tested.By analyzing the detection process and test results,the detection error of the system is 0.06 mm.It is obvious that the detection result is reasonable,Through theoretical analysis and testing experiments,the detection system and detection algorithm designed in this paper can meet the detection requirements of multi-circle workpiece.The improved edge detection algorithm can improve the effect of edge detection and improve the accuracy of circle detection results.It has a certain exploration significance for multi-circle workpiece detection based on machine vision.