Map data acquisition method, device, system and storage medium for congested road sections

Abstract

The present invention provides a method, device, system and storage medium for collecting map data of congested road sections. The method includes: collecting images and image trajectories of congested road sections in a first time period; collecting point clouds of congested road sections in a second time period and the point cloud trajectory; according to preset rules, find the point cloud trajectory point matching the image from the point cloud trajectory as the target trajectory point; according to the recording time of the target trajectory point, modify the time of the image, and the The time of the image trajectory point corresponding to the image is obtained to obtain the modified image and the modified image trajectory. In this way, the image and image trajectory of the congested road section are collected in the first time period, and the point cloud and point cloud trajectory of the congested road section are collected in the second time period, and the two parts of data are automatically matched and fused to obtain map data that meets the accuracy requirements. It solves the problem of map data quality caused by point cloud data occlusion of congested road sections, and improves the accuracy and efficiency of subsequent high-precision map production.

Description

technical field [0001] The invention relates to the technical field of high-precision maps, in particular to a method, device, system and storage medium for collecting map data on congested road sections. Background technique [0002] High-precision maps are an indispensable and important supporting technology for autonomous driving, and map data collection relying on multi-sensor integrated mobile collection vehicles is the core channel for high-precision map data acquisition, and it is also one of the important links in its map production. With the increasing demand for high-precision maps, this requires map data collection to provide more accurate and fresher raw data. [0003] At present, map data collection is generally carried out by multi-sensor integrated mobile collection vehicles for field data collection. For example, if a mobile collection vehicle is equipped with various measurement devices such as laser scanners, panoramic cameras, integrated navigation (inert...

AI Technical Summary

Problems solved by technology

[0004] However, such a map data collection method is more susceptible to the influence of the external complex environment, which reduces the quality and efficiency of data collection, and affects subsequent high-precision map production.

For example: during the driving process of the mobile acquisition vehicle, it is affected by surrounding vehicles and pedestrians, resulting in the loss of image or point cloud data

In the process of making high-precision maps, it is necessary to manually draw the missing parts, and the manual workload is very large

In particular, there are some road sections that are congested for a long time during the day, and the data loss of these road sections is even more serious

When the incomplete area is large, even manual work cannot map the missing parts that meet the accuracy requirements

Images