Research On Active Positioning Technology Of Sanitary Ware Casting Based On Machine Vision

The casting need drilling,surface cutting and other mechanical processing after forming.In machining process,it is necessary to limit and calibrate its position on the processing platform.Usually,there are two kinds of common ways: one is manual precise alignment,which is often difficult to achieve the required accuracy,and it takes a long time and a lot of manpower;the other is the production of precision fixture,which changes the production process conditions due to the changes in the precision of production equipment and mould.All of these will cause the change of products in the process of casting forming,to a certain extent,the precise fixture will lose its function.In order to overcome the shortcomings of the above two methods,an active positioning and alignment system based on machine vision was developed to realize the precise positioning and automatic alignment function of the sanitary ware casting.This paper mainly includes the following aspects:Firstly,aiming at the requirement of the project,the system structure of vision system and motion control system is studied and designed,and the selection of various hardware devices is carried out.Secondly,according to the analysis of the structure of a specific sanitary ware casting,three marker points for positioning are selected.The image preprocessing algorithms such as smoothing filter,threshold segmentation and edge extraction are studied.A set of image preprocessing schemes suitable for this topic is selected through programming experiments.Thirdly,in order to extract the accurate location information of landmarks,the basic principle of Hough circle detection method is analyzed firstly,and then the gradient Hough circle detection algorithm of OpenCV is used for feature extraction.The experiment shows that this method is not stable in the application of this subject and the effect of circle detection is poor.The source code of this method is analyzed,the evaluation strategy of the algorithm is improved,and the optimized algorithm is used for feature extraction.The circle detection experiments show that the algorithm meets the accuracy requirements of the subject.Fourthly,extracting the accurate coordinates of the marking points,the positioning algorithm combined with the motion control system was used to complete the active positioning operation of the sanitary ware casting,and the experimental results and error reasons were analyzed.