Laser SLAM positioning system and method based on deep learning

Abstract

The invention discloses a laser SLAM positioning system and method based on deep learning, and the system and the method comprises the modules and steps: 1, a point cloud preprocessing module which is used for extracting feature of ground points, planes and edges; 2, a laser speedometer module which is used for optimizing the pose in two steps, forming height constraint by ground points and optimizing the distance between the features; 3, a laser mapping module which is used for performing local optimization after down-sampling of adjacent frames, and optimizing mapping pose data in combination with odometer pose data; 4, a loopback detection module which is used for solving inter-frame similarity by using descriptors in a SegMatch framework and combining a twin neural network; and step 5, carrying out repositioning after the loopback is detected, and obtaining a corrected current pose after global optimization is carried out. Based on the method, the positioning precision of a mobile robot in indoor and outdoor environments is effectively improved.

Description

technical field [0001] The invention relates to the technical field of mobile robot positioning, in particular to a laser SLAM positioning system and method based on deep learning. Background technique [0002] With the rapid development of robot technology, more and more technological products such as unmanned vehicles and unmanned aerial vehicles have begun to enter people's lives, and an important technology to realize the intelligence of various mobile robots is positioning. In practical applications, when a mobile robot faces complex scenes, such as light changes and many dynamic obstacles, it is easy to cause tracking failure, which in turn affects the positioning and mapping process of the mobile robot. Therefore, Simultaneous Localization and Mapping (SLAM) in complex environments is a hotspot and an important direction of mobile robot research. [0003] Lidar is an active detection sensor that does not depend on external lighting conditions. It has high-precision d...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a laser SLAM positioning system and method based on deep learning. The method effectively solves the problem of high complexity of traditional filter-based laser SLAM through hardware such as an embedded computing platform, a multi-line laser radar, and a robot mobile platform. The positioning accuracy is not high. At the same time, the method uses loop detection based on deep learning to further reduce the cumulative positioning error. This method has high robustness, high precision, and strong applicability

Images