Map data evaluation device and system, data acquisition system, acquisition vehicle and acquisition base station

Abstract

The invention discloses a map data evaluation device and system, a data collection system, a collection vehicle, and a collection base station. The map data evaluation device includes: a data acquisition module, which is used to obtain real-time field data collected by an automatic driving map through a mobile network; analysis and judgment The module is connected with the data acquisition module, which is used to analyze and judge whether the data collected in the field of the automatic driving map is missing, and output the analysis results; the quality evaluation module is used to check the quality of the panoramic image, and output the inspection results; registration accuracy calculation The module is used to analyze the trajectory of the panoramic camera and the laser point cloud for the original data of the laser scanning and the original data of the panoramic camera, and detect the registration accuracy of the laser scanning data and the panoramic image data based on mutual information, and output the registration error. Therefore, by implementing the present invention, the quality of data collected in the field can be evaluated in time, and the collection quality and production efficiency of map data can be improved.

Description

technical field [0001] The invention relates to the field of map data collection, in particular to an automatic driving map data evaluation system, a map data evaluation device, an automatic driving map data collection system, a collection vehicle, and a collection base station. Background technique [0002] Autonomous driving maps are an indispensable and important supporting technology for autonomous driving, and field data collection relying on multi-sensor integrated mobile acquisition vehicles is the core channel for autonomous driving map data acquisition and one of the important links in its map production. With the rapid development of autonomous driving technology, high-precision maps are required to mature quickly to provide more accurate and vivid map data. [0003] The core channel for autonomous driving map data acquisition is to collect field data through multi-sensor integrated mobile collection vehicles, such as laser scanners, panoramic cameras, integrated n...

AI Technical Summary

Problems solved by technology

[0003] The core channel for autonomous driving map data acquisition is to collect field data through multi-sensor integrated mobile collection vehicles, such as laser scanners, panoramic cameras, integrated navigation (inertial navigation + GPS), odometer (DMI ) and other measurement equipment, such a complex field data acquisition system includes multiple modules such as combined sensors, mechanical structure, electronics and electrical, data communication storage, etc., which is more susceptible to the influence of the equipment itself, the acquisition personnel and the external complex environment. Reduce the quality and efficiency of its data collection

[0004] However, in the actual collection process, the inventors of the present invention found that the quality inspection of the existing field data collection for automatic driving maps is carried out after the data collection is completed. Due to the complicated data preprocessing process and other reasons, the data quality assessment and inspection are completed hours or even days after the field collection task is completed

It takes several hours or even a few days after the field collection task to complete the data quality assessment and inspection. If a quality problem is found at this time, it will cost a lot to re-collect or re-collect the data, which will seriously affect the efficiency of field data collection for autonomous driving maps. Larger waste of human and material resources

[0005] In addition, the inventors of the present invention also found that: the above-mentioned automatic driving map field data collection vehicle is not robust enough, and its data quality inspection timeliness is poor. Therefore, a real-time and effective data quality evaluation solution is urgently needed to realize the accuracy of the collected data. Real-time quality inspection and problem feedback

Images