Robot control method and system based on multi-line laser radar and event camera SLAM

Abstract

The invention provides a robot control method and system based on a multi-line laser radar and an event camera SLAM. The method comprises the steps: generating a 3D map according to image data collected by a binocular event camera; acquiring laser point cloud data acquired by a multi-line laser radar, wherein the multi-line laser radar and the binocular event camera are jointly calibrated and synchronized in advance; fusing the laser point cloud data and the 3D map to obtain a fused 3D map; performing 2D mapping on the fused 3D map to obtain a planar grid map; acquiring robot pose data measured by an inertial measurement unit; and generating an obstacle avoidance motion control instruction according to the planar grid map and the robot pose data, and sending the obstacle avoidance motion control instruction to a robot. Therefore, robot obstacle avoidance control can be realized stably and accurately at a high speed.

Description

technical field [0001] The present invention relates to the technical field of robot control, in particular to a robot control method and system based on multi-line laser radar and event camera SLAM. Background technique [0002] The robot control system is the "central brain" of the robot's movement, which can assist the robot to process external information, and output corresponding instructions to control the robot to respond accordingly. For self-moving robots, SLAM (Simultaneous Localization and Mapping, Synchronous Localization and Mapping) technology is generally used to realize the obstacle avoidance function, but the traditional control system can only move and respond at low speed, but it is powerless to track and avoid high-speed objects . [0003] At present, the general SLAM control system is equipped with traditional cameras or lidar sensors. Among them, traditional cameras have high requirements for environmental lighting, and there are cumulative errors in m...

AI Technical Summary

Problems solved by technology

For self-moving robots, SLAM (Simultaneous Localization and Mapping, Synchronous Localization and Mapping) technology is generally used to realize the obstacle avoidance function, but the traditional control system can only move and respond at low speed, but is powerless to track and avoid high-speed objects

[0003] At present, the general SLAM control system is equipped with traditional cameras or lidar sensors. Among them, traditional cameras have high requirements for environmental lighting, and there are cumulative errors in matching. Further processing is required to reduce the impact of errors, resulting in increased computing costs. Moreover, in When identifying and tracking high-speed moving objects, it is necessary to perform discrete sampling and perform morphological processing to extract feature points and achieve point cloud reconstruction through matching. The calculation cost is high and the response time is long, resulting in the movement of objects within the monitoring range beyond a certain limit. When the speed is high, there will be a problem of tracking loss; although the lidar has high accuracy, it still has problems such as limited detection range and the inability to distinguish semantics from the reconstructed map.

Therefore, the information collected by the traditional SLAM-based robot control system is greatly affected by the environment, and its stability and accuracy are limited, making it difficult to further improve

Images