An algorithm of robot SLAM based on self-detected waypoint is introduced to realize robot's mapping in its interesting area. With characteristics of large information in the interesting area and dense landmark, and clustering the landmark in the interesting area, guiding robot's movement with clustered central point as robot's next step waypoint is performed. Robot clusters the observed area in its observation every time and takes the clustered center based on the largest number of landmark as the robot's next step waypoint, and the basic framework of robot SLAM is estimated on EKF-SLAM. Simulation experiment shows, due to robot keeping movement toward the area with dense landmark, this method increases the number of robot's observed landmark and frequency of observation, enhances robot's positioning accuracy and realizes to detect its waypoint autonomously.
展开▼