Image Jacobian-Based Uncalibrated Visual Servoingfor Industrial Robot

Robot visual servoing is the fusion of results from many elemental areas including real-time image processing, robot kinematics, robot dynamics, control theory, computer technology and real-time computation, and is a main subject in the research field of computer vision.MOMOMAN-SV3XL industrial robot visual servoing control system is constructed, in which experiments of visual positioning and visual servoing tasks are conducted in this paper. The on-line estimation algorithm for Image Jacobian matrix model was designed based on the Kalman filter and Sage-Husa adaptive filter. With this Image Jacobian on-line estimator, the visual feedback controller is designed for the tasks of positioning and tracking by uncalibrated vision sensor. The main work in this paper is summarized as follows:1. Give a full discussion on the features of Image Jacobian matrix and its use in uncalibrated visual servoing system. Since the relation between the visual feedback space and the robot motion space is nonlinear, when using linear image Jacobian matrix to describe it, the image Jacobian matrix should be estimated iteratively on-line. For this task, construct a dynamic system with state variables formed from elements of the image Jacobian matrix. Thus Kalman filter and Sage-Husa adaptive filter are separately used to estimate the state variables of the constructed system on-line.2. Present a visual feedback control scheme for the task of positioning and tracking in the uncalibrated environment. The proportion and the proportion integral control are separately used for visual positioning and visual tracking in visual feedback. The Kalman filter and the Sage-Husa adaptive filter based on-lineestimation algorithm for Image Jacobian are separately used with finished movements, and avoid the effect of the redundant try movements.3. Construct robot visual servoing hardware system and program control software with VC++, carry out many experiments with the control strategy described above. The results show the validity of the algorithm, and the well performance of the corresponding visual positioning and tracking.