Research On Target Recognition And Location Of Industrial Robot Based On Computer Vision

With the rapid developments of computer technology, image processing technology and pattern recognition technology, machine vision is applied more widely. As one important way to obtain environment information, the machine vision can improve robotic flexibilities and autonomy. It is a key technology that how robots with vision to capture exact image information, as well as to extract the feature parameters of components real-timely, to recognize the component types, and to judge the position and posture of component. Based on the reason, the paper presents the algorithms of target recognition and position in industrial robot, especially real-time match recognition technology and space position technology in detail. The main task follows:First of all, the captured original image should be preprocessed in order to enhance image quality. The main preprocess ways are median filtering and histogram equalization. In the process of feature extraction, edge feature is used. After comparing Canny operator algorithm to other edge detection algorithms, Canny operator algorithm is introduced to extract edge features, which shows the effect of high positioning and response unicity.Secondly, based on traditional image match algorithm, an improved genetic algorithm and component recognition about Hausdorff distance is offered in this paper. The algorithm uses component edge feature as match feature, adopt corrected Hausdorff distance as the norm of target similarity measurement, introduce genetic algorithm to search best object to match. Then, the target match recognition is achieved in distance transformation space. The experiment results show that this method can detect the target or the shaded target in the motion of parallel displacement, rotation or microscale change efficiently.Thirdly, for component's space location, constant rotation array method is introduced to locate the 3D positioning and the depth information by moving monocular vision(camera amounting on robot's connecting-rod 3). The method is keeping the rotation array between robot's connecting-rod 3 and robot reference coordinate constant, which can simplify the complicated hand-eye calibration and camera calibration.Last, by using the recognition and position methods referred above and the hardware system, which consist of GRB-400 robot, CCD camera, image sampling card and PC, grabbing component experiment is accomplished.