In order to improve the palletizing efficiency of enterprise workpieces and reduce the labor intensity of workers, an industrial robot visual palletizing experimental system is built. Based on the analysis of the kinematics model of the industrial robot, the position and posture information of the end effector of the industrial robot is obtained through the coordinate relationship between the camera coordinate system and the end coordinate system. An image removal algorithm based on workpiece position coordinates is proposed to eliminate redundant workpiece images; the actual position of the workpiece to be palletized is calculated by the rotation and translation matrix, and this position is used as the position of the industrial robot to grasp the workpiece after vision correction. The experimental results show that the repeatability positioning accuracy of the improved system is significantly improved in X and Y directions, which verifies the accuracy of the industrial robot visual palletizing system model.