Application of Computer Vision Technology Based on Neural Network in Path Planning

Abstract

Path planning (PP) is a hot topic in the field of robotics, and neural networks (NNs) and computer vision technologies are widely used in the PP of robots. In this paper, the application of robot PP is classified, compared and analyzed, the autonomous dynamic obstacle avoidance of the robot in the dynamic obstacle environment is realized, and the particle swarm algorithm is used to find the global optimal path. In this paper, a fuzzy optimization algorithm for PP of visual robots is proposed. The algorithm is based on predictive control rolling optimization, and expresses the system optimization goals and constraints by membership, so as to realize the multi-objective optimization problem. Simulation experiments show that the proposed algorithm is effective. The experimental results show that compared with the traditional PP of the robot based on NN, the NN can find the shortest path for the robot in the dynamic avoidance of obstacles, which greatly optimizes the avoidance time. Therefore, it is very valuable to use NNs and computer vision technology to study PP.