Laser SLAM system and method based on feature extraction algorithm and key frame

Abstract

The invention relates to a laser SLAM system and method based on a feature extraction algorithm and a key frame. The laser SLAM system comprises a feature extraction module, a tracking module and a mapping module; wherein the feature extraction module is used for extracting plane feature points and edge feature points of the current frame; the tracking module is used for tracking the current frame in real time according to the plane feature points and the edge feature points of the current frame, determining the pose of the current frame and determining whether the current frame is a key frame; and the mapping module is used for matching a key frame to a local map when the current frame is the key frame, carrying out loopback detection according to the position of the key frame, optimizing the pose and obtaining a global consistent map. In the method, a more accurate pose and a global map are obtained by matching a key frame to a local map; in addition, a loopback detection thread is used for inhibiting accumulated errors of a large scene.

Description

technical field [0001] The present invention relates to the technical field of SLAM (Simultaneous Localization And Mapping, simultaneous positioning and map construction) in the field of robotics, in particular to a laser SLAM system and method based on feature extraction algorithms and key frames. Background technique [0002] Reliable and accurate simultaneous localization and mapping (SLAM) technology is very important for mobile robots and unmanned vehicles to perform environment perception, path planning and other tasks. SLAM has received extensive attention in both academia and industry. In the past few decades, vision-based SLAM methods and laser-based SLAM methods have achieved good results on robots. Vision-based methods have a large viewing angle and rich texture features, but the lack of depth information and sensitivity to illumination make it difficult for perspective SLAM to land in outdoor scenes. Compared with vision, laser has direct and reliable depth mea...

AI Technical Summary

Problems solved by technology

The more representative systems are LOAM and LeGo-LOAM. They also achieve good results in many scenes, but their feature extraction has no perspective invariance and distance invariance, which is not robust enough; and the calculation efficiency, accuracy and Robustness is also not well balanced

Images