Road positioning mark serving unmanned vehicle

Abstract

The invention relates to a road positioning mark serving an unmanned vehicle, a color block array with a simple geometrical shape is used for representing unique and accurate road geographic position information, accurate positioning information is provided for the unmanned vehicle, data bits are represented through the color block array, and camera recognition and image processing are easy; the geographic position information represented by the road positioning mark provided by the invention is unique geographic information based on global positioning, and is more applicable compared with a road identifier based on a relative geographic position in the prior art; compared with a standing type or suspension type road positioning mark in the prior art, the road positioning mark provided by the invention is small in size, can be assembled and laid on the road ground, is small in occupied area, facilitates image acquisition and road construction, and can reduce the production and installation cost.

Description

technical field [0001] The invention belongs to the technical field of automatic driving positioning, and in particular relates to a road positioning mark serving unmanned vehicles. Background technique [0002] Self-driving cars can perceive the external environment and drive safely without human intervention. Its reliability and safety are related to whether it can obtain accurate positioning information. Positioning technology is already a key part of the control of automatic driving systems. [0003] The positioning accuracy of traditional GPS positioning technology can reach 16 feet in open areas, but in densely built areas, tunnels, three-dimensional traffic systems such as viaducts, etc., the accuracy of GPS drops to 10-20 meters, and it cannot provide height direction. Location information, such technical insufficiency will affect the efficiency and safety of traffic flow, and limit the application in the field of automatic driving. [0004] In addition to satellite...

AI Technical Summary

Problems solved by technology

[0003] The positioning accuracy of traditional GPS positioning technology can reach 16 feet in open areas, but in densely built areas, tunnels, three-dimensional traffic systems such as viaducts, etc., the accuracy of GPS drops to 10-20 meters, and it cannot provide height direction. Position information, such technical insufficiency will affect traffic efficiency and safety, and limit the application in the field of autonomous driving

[0004] In addition to satellite-based positioning, the current mature driving positioning technology also includes RFID-based wireless positioning. However, due to the high cost, this technology has not been widely popularized. Currently, it is mostly used in rail transit and logistics parks, etc. traffic system

[0005] Other positioning technologies such as Cell ID, AFLT (forward link trilateration), AGPS (assisted GPS), RTK (real-time dynamic carrier phase difference technology), etc., can solve the problem of GPS positioning accuracy to a certain extent, but they are still not very good It can effectively solve the signal reflection and occlusion in the urban environment, the sensitivity is low, and the requirements for the technical environment are high

[0006] At present, in the vehicle-road coordination system, especially in the field of automatic driving, relevant research has designed a white circular color block mark on the road surface, using the imaging device installed on the vehicle to collect images, and using the weighted least square method to simulate the image. A vehicle precise positioning method combined with positioning information analysis, this method only uses simple dots and spaces to express binary codes, the efficiency of expressing information is not high, but takes up a lot of road space, and the location information it carries is not It is not the only accurate three-dimensional geographic location information based on global positioning, but only the relative position data of each logo. When the client embeds it and develops it, it needs to load the matching database, which is not conducive to the development of offline client devices.

In addition, this method also requires precise calibration points, and its positioning effect is not good when driving at night or in environments with poor lighting conditions.

Images