Method and device for generating a point cloud map, and a computer system

Abstract

Provided are a method and device for generating a point cloud map, and a computer system. The method comprises: initializing the spatial parameter of each pixel in a first picture, wherein the first picture is any one of a two-dimensional picture set, the spatial parameter includes at least the depths of three-dimensional spatial points corresponding to the pixels; updating the spatial parameter of each pixel in the first picture according to a matching error function, wherein the matching error function is associated with the errors of the three-dimensional spatial points corresponding to the pixels; determining, based on the updated spatial parameter of each pixel in the first picture, a depth map corresponding to thefirst picture; generating a dense point cloud map according to the depth map corresponding to the first picture. The method and device for generating the point cloud map and the computer system are capable of efficiently generating a dense point cloud map.

Description

[0001] Copyright statement [0002] The disclosure of this patent document contains material that is protected by copyright. This copyright belongs to the copyright owner. The copyright owner has no objection to the reproduction by anyone of the patent document or the patent disclosure as it exists in the official records and files of the Patent and Trademark Office. technical field [0003] The present invention relates to the field of information technology, and more specifically, to a method, computer system and device for generating point cloud maps. Background technique [0004] The existing map construction generally adopts synchronous positioning and map construction technology, that is, the robot locates its own position and posture through the map features repeatedly observed during the movement, and then constructs the map according to its own position and posture, so as to achieve simultaneous positioning and map construction. Purpose. Under this idea, the two ...

AI Technical Summary

Problems solved by technology

The maps generated by this method are all in the form of point clouds, which are generally just a rough outline and not fine enough, that is, only sparse or semi-dense point cloud maps can be generated.

[0005] However, the above-mentioned non-dense map is likely to miss important features in navigation, such as missing traffic lights, or cannot provide accurate information on whether it is allowed to pass, so the above-mentioned scheme is not suitable for generating more accurate maps or generating more accurate maps. Realistic Vision Modeling Tasks

Images