Application of Automatic Recognition and Positioning Technology of Industrial Robot based on Monocular Vision Technology in Vehicle Manufacturing

Abstract

Combining machine vision technology with intelligent algorithms to improve the automatic identification and positioning technology of industrial robots to meet the production needs in different industrial environments is the current research direction of intelligent robots. In this study, an automatic identification and positioning technology for industrial robots based on monocular vision technology is proposed. First, the template matching algorithm and the Scale-invariant feature transform (SIFT) algorithm are introduced. Aiming at the shortcomings of the Hessian matrix in the SIFT algorithm in the process of eliminating boundary effects, an improved SIFT algorithm using Harris corner point detection is further proposed, and the improved SIFT algorithm is used in the automatic recognition and target positioning operations of industrial robots. In order to verify the performance of the proposed improved SIFT algorithm, the recognition accuracy and positioning angle deflection of the algorithm in different plastic sheets were detected. The experimental results show that under the improved SIFT algorithm, the recognition accuracy of five different styles of plastic sheets is above 98%. The improved SIFT algorithm also has less error between the predicted value and the actual value of the positioning angle deflection on the four plastic sheets. The robot under the improved algorithm is applied to the vehicle manufacturing industry and the production efficiency of the vehicle is improved through the automatic recognition and positioning technology of intelligent robot.