Research On Multi-hole Positioning System Of Panel Light Based On Machine Vision

Screw locking is an important technology in panel light production.It is necessary to accurately identify the each hole position of thread in order to avoid leakage and floating lock.Traditionally,the location of threaded holes is realized by human eye recognition or by fixture.But it is difficult to apply to rapid industrial production due to the low accuracy and efficiency of positioning.In order to solve the above problems,an effective panel light positioning system based on machine vision is studied in this paper.The main contents of this paper are as follows:1.Considering the requirement of speed and precision for panel light lock,the detection scheme is formulated.appropriate hardware and software are selected to build the experimental platform.2.In order to locate the threaded hole accurately,considering the effect of lens distortion on the measurement hole position,the image distortion correction is realized by LabVIEW software,and the image measurement errors before and after distortion correction are verified by point distance experiment.3.Based on NI vision,the original image is preprocessed,and the pixel coordinates of the threaded holes on the lamp are obtained by designing a matching program through LabVIEW software.The coordinates of the holes are converted to the coordinates of the robot by combining the robot hand-eye calibration.4.Modbus serial communication is used to realize the transmission of the center position of the threaded hole.The hole coordinate is transmitted to SCARA robot,and the control program is designed to realize screw locking.Through the debugging and running of the actual instrument,the effect of locking is verified.Through theoretical analysis and experimental testing,the research in this paper is based on machine vision,which improves the positioning accuracy and meets the real-time requirements,and can be applied in the production of panel light screw lock.