Point cloud map construction method and device, computer equipment and storage medium

Abstract

The invention relates to a point cloud map construction method and device, computer equipment and a storage medium. The method comprises: obtaining point cloud data and corresponding vehicle body sensor data, performing fusion processing on the vehicle body sensor data to obtain an optimal estimation state of the vehicle body sensor data and corresponding uncertainty information, and constructinga point cloud map corresponding to the point cloud data according to the optimal estimation state and the corresponding uncertainty information. By adopting the method, the states of all the point cloud data can be integrally optimized according to uncertainty information, the uncertainty of all the states, especially the uncertainty of the state lacking a position observation scene (such as a tunnel scene) for a long time, is reduced, and the overall positioning precision is improved; and the high-precision map construction error caused by lack of position observation information for a long time and insufficient positioning precision is remarkably reduced.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a method, device, computer equipment and storage medium for constructing a point cloud map. Background technique [0002] High-precision map technology can be applied in the field of mobile robots and autonomous driving. For example, with sensors such as LiDAR and cameras, the centimeter-level positioning of mobile robots can be realized. [0003] The existing methods for constructing high-precision maps are mainly based on nonlinear optimization methods, constructing a cost function through the state of the sensor, and then optimizing the cost function to obtain a high-precision map. In scenarios such as long tunnels, due to the large uncertainty of the initial value, the nonlinear optimization tends to converge to a wrong local solution, resulting in large errors in the constructed high-precision map. Contents of the invention [0004] Based on this, it is necess...

AI Technical Summary

Problems solved by technology

[0003]Existing methods for constructing high-precision maps are mainly based on nonlinear optimization methods, constructing a cost function through the state of the sensor, and then optimizing the cost function to obtain a high-precision map, but , this method lacks position sensor observation for a long time and the scene with inconspicuous features, such as in a long tunnel, due to the large uncertainty of the initial value, the nonlinear optimization tends to converge to the wrong local solution, thus making the constructed high There is a large error in the fine map

Images