Robot repositioning method, device and equipment and storage medium

Abstract

The invention discloses a robot repositioning method, device and equipment, and a storage medium. The method comprises the steps of obtaining an image key frame captured by a photographing device of a robot, wherein the image key frame comprises a 2D image; performing image matching on the 2D image and a map key frame pre-stored by a robot to obtain an estimated pose; performing point cloud transformation on the image key frame to obtain a first 3D point cloud, and obtaining a second 3D point cloud of the map key frame, wherein one 3D feature point of the first 3D point cloud and the corresponding 3D feature point of the second 3D point cloud serve as a matching point pair, and each pair of matching points corresponds to the same point in a real scene; calculating a distance error between each pair of matching points, and determining a total distance error according to the distance errors; obtaining a repositioning pose according to the total distance error and the estimated pose; carrying out relocation by combining the 2D image and the depth image, effectively correcting the wrong 2D relocation pose, and improving the relocation precision.

Description

technical field [0001] The invention relates to the technical field of robot positioning, in particular to a robot relocation method, device, equipment and storage medium. Background technique [0002] In recent years, the relocalization problem has become a hot topic in the field of mobile robotics. Relocation can help the robot to correct its accumulated errors and find its own correct position when it loses its positioning. The relocalization of mobile robots is a basis for mobile robots to realize intelligent navigation and environment exploration, and it is also one of the key technologies for mobile robots to realize real intelligence and complete autonomy. At present, the relocalization of mobile robots is divided into vision-based relocalization and laser-based relocalization. The laser-based relocation technology uses laser radar, which has higher accuracy, but the cost of using laser radar is much higher than that of cameras, which is not suitable for widespread ...

AI Technical Summary

Problems solved by technology

The laser-based relocation technology uses laser radar, which has higher accuracy, but the cost of using laser radar is much higher than that of cameras, which is not suitable for widespread use. Therefore, the market mainly uses camera-based vision-based relocation technology. , but the vision-based relocation technology mainly uses the monocular vision scheme, and the positioning accuracy is relatively low.

Images