Research On Visual Measurement System For Position Size Of Multi-Round Hole Parts

Round hole is an important feature in machined components.As a common mechanical part with round hole characteristics,hole-group parts are widely used in various mechanical structures.The detection of parts with round hole has become one important content of their inspections.The traditional measurement of the round hole size is mainly carried out by manual mechanical measuring tools,which not only greatly affects the detection efficiency,but also is not conducive to its automation and integration.The use of digital and non-contact vision measurement technology can effectively improve these defects.In this paper,a visual measurement method for the size of circular holes and the position between circular holes of a typical hole group part is studied and verified by experiments.The main research contents of this paper include:(1)For the specific measurement scene,the overall design of the system hardware and its selection are completed,and the vision system is calibrated.After analyzing the requirements of system function,the human-machine interface of the system is designed.(2)Due to the existence of the lens distortion and the intersection angle between the imaging plane and the workpiece plane,the circular feature of the holes mutate into ovals or other shapes in the image,which increases the complexity of the feature detection algorithm and reduces the detection speed and accuracy.After analyzing the causes of the image mutation,this paper puts forward one kind of image correction method by angle and distortion correction to reflect the actual workpieces.(3)In order to extract the target edge,firstly,the original image is preprocessed by hybrid filtering algorithm and sharpening the diffuse edges.Then,by analyzing the characteristics of the surface of the workpiece,the object edges of the workpiece are extracted by using the decomposition method based on the improved LP image pyramid.Compared with the traditional edge detection algorithm,this method can effectively eliminate the false edges generated by the fine grains and other features on the workpiece surfaces.In order to improve the detection accuracy of rough edges,sub-pixel detection technology based on multiple interpolation is used to obtain fine edges.(4)After analyzing the applicability and inherent defects of Hough transform,it is improved to speed up the algorithm.The parameters of circular feature are obtained by fast random Hough transform of edges,and then the distribution positions and sizes of circular hole are obtained.The reliability and accuracy of the system are verified by a large number of experiments.