Research On Rapid Edge Recognition Technology For Turning Aluminum Alloy Die Casting Parts

Aluminum die casting working surface and finishing surface need to be turned,and the process often produces cutting burrs on the edge of the surface,which affects the quality of the product and requires manual deburring.Manual deburring has the problems of high labor cost and labor intensity,so it is necessary to use robots to replace manual deburring.Currently,off-line programming based on digital model is often used to allow the robot to drive the tool to follow a fixed trajectory for deburring.However,due to the difference in deformation of each aluminum alloy die casting during the production process,there is a large gap between the edge curve of the digital die and the actual edge,which makes it impossible to perform burr removal accurately.Therefore,it is necessary to use the machine vision method for accurate identification and extraction of contour edge curves to guide the robot for burr removal of aluminum alloy die castings.The main research contents of the paper are as follows:(1)Aluminum alloy die casting model identification and edge extraction are completed.Based on the shape template matching algorithm with improved similarity measure,the local edge templates of different aluminum alloy die castings are established,and the local edge templates are used to identify the aluminum alloy die casting models and extract the coordinates of the edge profiles.(2)A deformation template matching algorithm based on extracting the extension of edge coordinate points is proposed.Due to various processing errors in the machining process of aluminum alloy die casting,rigid template matching does not adapt well to the variation of aluminum alloy die casting contour differences.The actual edge coordinate points are extracted using the extracted edge coordinate point extension region to achieve the deformation of the matching template and meet the actual production requirements.(3)A height-based coordinate error compensation model is established.When extracting the contour edge coordinate points from 2D images,the height of different die castings is different and there are also height differences on the surface of die castings,which leads to errors in extracting the edge points.By calculating the error effect of different heights,the height error compensation relationship is established to reduce the height error effect and ensure the accuracy of the edge coordinate points.(4)The spatial contour edge extraction based on the point cloud data of aluminum alloy die casting is realized.The structured light camera is used to scan the aluminum alloy die casting and obtain the 3D point cloud of the aluminum alloy die casting.The point cloud data is filtered to obtain the required point cloud data set,and the 3D point cloud data is projected to the 2D plane,and the edges of aluminum alloy die casting are extracted in the 2D plane and then mapped into 3D coordinate points to realize the extraction of aluminum alloy die casting edge data from the 3D point cloud data of aluminum alloy die casting.