Research on Obstacle Avoidance in Automatic Driving Based on SLAM Path Planning

Automatic driving based on SLAM is known as one of the key research topics today. At present, researchers have made progress in some aspects such as EKF-SLAM and Graph SLAM and some even have commercialized attempts. However, the problem of automatic driving obstacle avoidance based on SLAM still lacks a unified cognition. Therefore, the research topic of this paper is the method of obstacle avoidance in automatic driving based on SLAM trajectory planning. This paper summarizes and discusses the relevant research results by integrating relevant literature and further expounds the application of SLAM technology in automatic driving obstacle avoidance: The front-end vision based on feature point method is more mature than other front-end tracking options. EKF-SLAM algorithm is not suitable for complex large-scale maps. Image optimization SLAM is dedicated to obtaining the optimal pose, but it needs to be initialized according to the environment. Through the summary and discussion of the research results, this paper not only shows the application value of SLAM in automatic driving obstacle avoidance, but also gives readers a reference for the choice of SLAM technology