Map and V2V data fusion model, method, system and medium

Abstract

The invention provides a map and V2V data fusion model. The map and V2V data fusion model comprises a roadside information extraction module, a vehicle end information extraction module and an environment information sending module, the roadside information extraction module transmits a real-time road condition image collected by a roadside camera sensor to an OBU carried on an intelligent driving vehicle through a V2I communication technology, detects whether pedestrians exist on the two sides of a lane or not through a quicker RCNN pedestrian detection algorithm, and marks in a dynamic layer of a high-precision map; the vehicle end information extraction module receives an alarm signal sent by a nearby intelligent vehicle through a V2V communication technology, and marks the alarm signal to a dynamic layer of a high-precision map; the environment information sending module sends the road condition detected by the intelligent driving vehicle to the outside through the V2V communication technology. According to the invention, a high-precision map is taken as a reference, and environment information from the roadside camera and other intelligent driving vehicles is fused, so that the intelligent driving vehicles can obtain detailed road environment information, and complete digitization of a test field can be realized.

Description

technical field [0001] The present invention relates to the technical field of unmanned high-precision mapping, environment perception, and sensor data fusion, and specifically relates to a map and V2V data fusion model, method, system, and medium. Background technique [0002] The traditional map abstracts the road network into a road-level topological relationship, that is, the road is used as the basic unit, and the intersection is used as the connection point to connect different roads, and the entire road network is formed through layer-by-layer association. Since human drivers can rely on strong environmental perception and logic processing capabilities, human drivers can use the navigation functions provided by traditional electronic maps to determine specific driving behaviors based on their own perception and logic processing capabilities. The "driver" of an intelligent driving vehicle is equipped with software systems and various sensors, and its environmental perc...

AI Technical Summary

Problems solved by technology

The "driver" of an intelligent driving vehicle is equipped with software systems and various sensors, and its environmental perception and logic processing capabilities are very weak. It is difficult for intelligent vehicles to achieve automatic driving entirely by their own processing capabilities. Even if it is realized, it can only be limited to some specific scene

Images