Design And Implementation Of Two-degree-of-freedom PTZ Visual Servo Control System

Recently,the PTZ equipped with vision system has played an important role in the fields of security monitoring,human-computer interaction and other fields.The camera is installed on a multi-freedom head,and the PTZ's flexible degrees of freedom are used to drive the camera to achieve fast target tracking.The PTZ vision servo control is based on the information obtained by the vision sensor to the servo controller to achieve closed-loop control of the gimbal,so that the target always appears in the center of the camera.The PTZ vision servo technology mainly involves disciplines such as control theory,digital image processing,etc.,which is a hot spot in the field of robotics.In this thesis,the two-degree-of-freedom PTZ visual servo of the mechanical hardware,controller design,and tracking algorithm design are discussed and studied.The main purpose of tracking is to improve the tracking control accuracy of the PTZ vision servo.This thesis adopt a double-loop structure control scheme,design and improve the position loop controller as the inner loop in detail.At the same time,the vision tracking algorithm is designed to ensure the target position is obtained steadily.Finally,the reliability of the controller and tracking algorithm in this thesis is verified on the physical platform.The experimental results of multiple groups show that the design method of this thesis can greatly improve the performance of visual servoing and fully meet the requirements for the use of small PTZ visual servo.The research content of this thesis is mainly reflected in two aspects of servo controller and vision tracking algorithm.During the design of the position loop,a linear quadratic regulator(LQR)based on internal model control was designed,and then an improved particle swarm algorithm was used to optimize the parameter selection of the linear quadratic regulator(LQR)weight matrix.The improved particle swarm optimization algorithm uses crossover operators and dynamic parameters to achieve cross-dynamic optimization of particles.In the process of target tracking,the traditional Cam Shift algorithm is used to obtain the position information during the movement of the target.Due to the problems of occlusion and environmental interference during the movement of the target,a certain prediction and tracking effect is achieved by combining with Kalman filtering.