Research On Control System Of LED Placement Machine Based On Machine Vision

Machine vision detection technology has the characteristics of fast,accurate and long-term operation.It has been widely used in various fields such as industry,medical care,life and other fields.With the surge in demand in the LED market and the expansion of the LED industry,imported general-purpose placement machines and traditional mechanical alignment placement machines have been unable to meet the production needs of LED products.Taking advantage of machine vision detection technology,it is of great significance to study the application of machine vision in the control system of LED placement machine.In view of the placement accuracy of the placement machine,this paper studies the control system of the LED placement machine based on machine vision.Firstly,the motion of each axis and the hardware structure of the LED placement machine are analyzed,and the overall design scheme is put forward.The key components of the placement machine are selected.The hardware connection of the control system is introduced,and the hardware platform of the LED placement machine is built.To study the application of machine vision technology in LED pose recognition,design a vision system that improves LED patch position and angle detection and recognition accuracy.This paper introduces the principle and method of LED image acquisition.The dual threshold segmentation method is used to identify the positive and negative polarity of LED.The minimum external rectangle method and the connected domain centroid fitting method are used to identify the pose of the LED respectively,and the algorithm test is carried out.The experimental results show that The connected domain centroid fitting method has higher recognition accuracy.In the QT5.0 development framework,the OpenCV open source library and motion control instruction library are implanted,and the LED placement machine software platform is developed using the C++ language.Through the demand analysis of the control system software,the overall design of the control system software design is designed,and the human-computer interaction interface,motion control layer,visual system layer and data storage layer of the control system are designed separately.The conventional PID control algorithm and the speed acceleration feedforward PID control algorithm are analyzed,and the PID parameters are set.The LED placement machine experimental platform is tested.The results show that the functions of the control system function well,and the placement positioning accuracy can reach 0.08 mm,and angle accuracy is up to 0.5°.