A hybrid localization method for mobile robots based on laser and vision

Abstract

The present invention discloses a laser and vision-based hybrid location method for a mobile robot. The mobile robot comprises a laser radar and a vision sensor. According to the technical scheme of the invention, the weight of each particle at a predicted position is updated based on the collected data of the laser radar and the collected data of the vision sensor. After that, particles of higher weights are re-sampled, so that the real location distribution of the mobile robot at the moment t can be obtained. Compared with the prior art, the above technical scheme integrates the high accuracy of the laser radar with the information integrity of the vision sensor, thus being wider in application range. Meanwhile, the defect that the visual location is unstable is overcome. In addition, in one embodiment of the present invention, a conventional particle filtering sampling model is improved, so that the diversity of particles is ensured.

Description

technical field [0001] The invention relates to a positioning method of a mobile robot, in particular to a hybrid positioning method of a mobile robot based on laser and vision. Background technique [0002] With the continuous development of mobile robots, simultaneous localization and mapping (SLAM) has become a major proposition in the development of mobile robots. How to solve this problem, there are many methods, from the model, it is divided into probabilistic positioning model (such as particle filter) and absolute positioning model (such as GPS), etc.; from the map, there are grid maps, topological maps, etc.; from the sensor The angle is divided into ultrasonic positioning, lidar positioning, visual positioning, etc. [0003] Probabilistic model is limited by memory and CPU, and absolute positioning model requires some equipment such as artificial beacons, or some special environments, such as indoors, which are not suitable for GPS; grid maps will have grid errors...

AI Technical Summary

Problems solved by technology

[0003] Probabilistic models are limited by memory and CPU. Absolute positioning models require some equipment such as: artificial beacons, or some special environments, such as indoors, are not suitable for GPS; there will be grid errors in grid maps, and the maintenance cost of topological maps is relatively high. It is difficult to recover afterwards; ultrasound has accuracy problems; the current normal laser radar is a two-dimensional laser radar, which has a low dimension and a low degree of recognition of the environment; visual positioning mainly includes feature-based positioning, positioning based on three-dimensional map matching, and positioning based on The positioning of three-dimensional point clusters, etc., the visual positioning algorithm is not very mature at present

Images