Vision-based robot obstacle avoidance method

Abstract

The invention discloses a vision-based robot obstacle avoidance method, which is characterized by comprising the following steps of: constructing a global positioning environment library; processing the visual image and the laser information of the current position of the robot, extracting feature information in the visual image and the laser information, performing retrieval and matching in the global positioning environment library, determining a real-time pose and a corresponding visual feature map of the robot, and performing global positioning on the visual positioning map to generate a local cost map; and adopting a depth vision sensor to obtain point cloud data containing obstacle information, subjecting the point cloud data to sparse processing, extracting key effective data, analyzing the extracted data, acquiring obstacle information in the three-dimensional direction in combination with laser data, and marking obstacle position information in a local cost map. According to the vision-based robot obstacle avoidance method, the functions of global positioning and three-dimensional obstacle avoidance can be realized at the same time, and the method has the advantages of small occupied area, light operation, high performance, low power consumption and the like.

Description

technical field [0001] The invention relates to the technical field of robot obstacle avoidance, in particular to a vision-based robot obstacle avoidance method. Background technique [0002] The technology of autonomous navigation and positioning is one of the key technologies of intelligent robots. The positioning technology can make the robot know "where I am"; the mapping technology can map the surrounding environment of the robot. It serves for positioning and also enables the robot to know "where I am". Where am I going”; navigation technology can let the robot know “how to get there”. In the case of an unknown environment, that is, without a map, the robot is required to have the ability to locate and map at the same time (SLAM, Simultaneous Localization And Mapping). In the case of a known environment (with a map), the robot needs to locate on the existing map and maintain global consistency. The sensors used for positioning and mapping are mainly cameras and lidar...

AI Technical Summary

Problems solved by technology

[0003] However, the current vision-related robot edge devices usually have the following problems: 1. For devices with vision as the core sensor, since the collected data must be processed based on the visual method of vision itself, the amount of calculation is large; In severe cases, it will not work normally

2. Usually, this type of equipment is responsible for one of the robot positioning function and obstacle avoidance function alone. It will waste a lot of space to install it on a small robot with a limited installation area, and there is also the problem of high power consumption.

3. The computing modules used by current vision-related edge devices usually have disadvantages such as low performance and high power consumption, and can only be equipped with relatively cumbersome mainstream Linux distributions

Images