Robot navigation method based on global map and robot using navigation method to navigate

Abstract

The invention discloses a navigation method based on a global map and a robot using the navigation method to navigate. The navigation method comprises the following steps: S1, enabling the robot to shoot videos of a ceiling and a wall in a room, and extracting the image distortion feature of each frame of image in the videos; S2, using an image matching method based on the contents, utilizing theimage distortion feature to match each frame of frame of the shot videos, extracting a key frame sequence according to the matching result, overlapping and connecting the key frames at the adjacent positions, and building the global map; S3, according to the image matching method based on the contents, matching the real-time video image of the robot and the key frames in the global map, finding the key frame which is most similar to the current vision image of the robot, and solving the global position of the robot in real time, so as to realize the navigation function of the robot. The navigation method has the advantages that the autonomous navigation positioning function of the robot can be realized, and the problem of easiness in kidnapping,and interference between similar objects andthe like in the navigation process of the robot can be effectively solved.

Description

technical field [0001] The invention relates to the fields of robot navigation and positioning, and more specifically, to a navigation method based on a global map and a robot using the method for navigation. Background technique [0002] An ideal indoor service robot should be able to autonomously plan walking routes, autonomously navigate and locate within the entire building, accurately shuttle between multiple rooms and corridors, and provide various services for humans. The prerequisite for this to be possible is for the robot to store a complete indoor map, preferably built by the robot itself. [0003] For this field, there are mainly two kinds of local map positioning and global map positioning. Local map positioning is mainly to construct maps based on indoor feature information, such as Simultaneous Localization And Mapping (SLAM). Among them, visual V-SLAM is the most widely used. The latest research includes ORB SLAM, dense SLAM, semi-dense SLAM, LSD SLAM, and ...

AI Technical Summary

Problems solved by technology

When the robot spins, moves sideways or slips, it cannot accurately measure its own spatial movement, and the extracted keyframes have a large error

At present, most of the key frame detection methods based on the change of frame content still only use the color statistical characteristics and feature point information of the object, and do not fully apply the layout and shape information of the object. Distortion, interference with content information

Images