Research And Development Of Robot Screw Lock Positioning And Detection System For TV Backplane Based On Machine Vision

With the increase of Chinese semiconductor display industry in the global market share,followed by the ability of TV monitors to boost demand,while the chain in the TV assembly production is still a lot to rely on the manual.Due to the increase in labor costs and the low stability of manual lock,automatic screw lock equipment is becoming more and more important for TV manufacturers.For TV backplane screw automation lock requirements,this paper adopts machine vision as the basis,6-axis robot as the sports platform,research and development of TV backplane robot screw lock bit and detection.The system has two functions: First,the TV backplane need to lock the screw hole by the visual positioning,access to the screw hole coordinates,so that the robot movement to the coordinates of the location to lock the screws;Second,locking screw quality by visual inspection,access to the screw image,to determine the quality of the lock.The main contents are as follows:First of all,this paper optimizes the traditional camera calibration algorithm,making it suitable for use in high-resolution cameras,and designs an efficient robot sensor calibration method based on the origin mark.Second,this paper studied the key technology of visual positioning of TV backplane screw holes,the hardware and software of the positioning system are designed,and the positioning method of the fine hole in the large field of view is studied.Taking into account the requirements of the screw lock accuracy,a quadratic positioning method is put forwards,through the thick and fine two industrial cameras to determine the TV backplane screw hole coordinate information.Coarse positioning through the TV backplane local image to determine the approximate range of the hole,fine positioning using the camera's small field of view one by one to find these areas within the precise coordinates,Locate one and lock one,with a high precision but low detection efficiency.Third,in order to achieve a higher positioning speed,the single positioning algorithm is studied using an industrial camera to shoot part of the TV backplane,getting part of the screw hole coordinates of the data,relying on the position between the screw hole to calculate the entire TV back Plate screw hole coordinates,but the accuracy of calculated screw holes coordinate of outside of view field cannot achieve the requirements of the screw lock.Fourth,optimized the single positioning algorithm,using two industrial cameras at the same time on the detection area for a shot,both to ensure high-speed positioning and avoid the coordinates of outside the field of view.The positioning speed in this way is fast,although the accuracy is lower than the secondary positioning method,at the same time it achieved the lock requirements.Finally,this paper studies the method of judging the quality of screw lock,designed an algorithm by the screw head cross slot to determine whether there are floating lock or missing lock algorithm.The research results of this paper are not only limited to the screw lock and detection of the TV backplane,in which the function of visual positioning can be extended to the positioning of the small size feature of the large workpiece in other fields,and the screw lock quality inspection function can also check the quality of the screw on the other surface of the workpiece.