Design Of Visual Guided Grab System For Copper Castings

As people's requirements for manufacturing continue to increase,it is difficult for industrial robots to complete complex tasks independently.The introduction of visual guidance technology can effectively overcome the limitations of robot operations.How to achieve rapid target recognition and precise positioning in visual guidance tasks is a difficult point,and also a key technology for visual guidance.In this paper,a template matching algorithm based on edge features is designed.A visual guided grabbing system is developed,and the visual guided grabbing task of copper castings is completed.The main research contents are as follows:(1)Research on the template matching method in vision guidance technology,deduce the principle of template matching method base on gray and geometric features,summarize the advantages and disadvantages of each method and the research direction.(2)On the basis of the template matching algorithm based on edge gradient feature,an exit method for feature point sampling matching is designed to improve the speed of template matching and anti-occlusion performance.The method firstly sorts the feature points of the template according to the spatial position,and then samples the sorted feature points to obtain a plurality of feature point sets.Finally,template matching is performed on the feature point sets.If the average similarity of a set match does not satisfy the minimum similarity threshold requirement,the match is exited,otherwise the feature points of all sets are matched.(3)A template matching experiment was designed to test the algorithm identification accuracy,anti-light variation,anti-noise and interference performance and positioning accuracy.The result shows that one hundred images containing the target are identified and positioned,and a total of eleven image recognition positioning results are unqualified;the algorithm can adapt to the effects of light,noise and interference;the position matchingaccuracy of the algorithm is 2 pixel,the angle matching accuracy is ±0.4o,and the average matching time is 150 ms.(4)The visual guiding and grasping system design scheme of the bathroom copper casting are determined,and the hardware platform of the visual guiding and grasping system is built.A set of open-sighted,full-featured and UI-friendly visual guidance software is designed to integrate the template matching algorithm in this software.The vision guidance system calibration of the robot "eyes outside the hand" is completed,and the visual guiding and grasping function of the bathroom copper casting is realized.The result showed that the recognition accuracy of the system is89%,the positional positioning accuracy is ±0.4263 mm,and the angular positioning accuracy is ±0.56°.