Automatic driving auxiliary safety early warning system in dark scene and early warning method thereof

Abstract

The invention relates to an automatic driving auxiliary safety early warning system in a dark scene and an early warning method thereof. The early warning system is mainly provided with a far infraredcamera, a temperature-sensitive camera, an illumination sensor, an alarm device, an inertial navigation unit, a control module and the like; through the fusion use of the sensors, a detection targetis subjected to self-defined semantic segmentation according to lane lines, running vehicles, pedestrians, road obstacles and other irrelevant objects, and the calculation efficiency of the system isimproved. The main functions are as follows: 1, automatic following driving, front target detection, anti-collision early warning, driving intention interpretation and pedestrian detection and anti-collision early warning in a dark scene are achieved; and 2, detecting, identifying and anti-collision early warning are carried out on unknown vehicles coming from the lateral blind area. The system can make up for the defect of low sensitivity of the visible light camera when the light intensity changes sharply in a dark scene, can detect unknown vehicles coming from a lateral blind area in the dark scene, gives out a safety early warning, and reduces the traffic accident rate.

Description

technical field [0001] The invention belongs to the technical field of automatic driving safety assistance, in particular to an automatic driving auxiliary safety early warning system in a dark scene and an early warning method thereof. Background technique [0002] With the increasing variety of sensor types and the diversification of artificial intelligence algorithms, autonomous driving technology is showing a trend of diversified development, but it also has higher technical requirements for existing computer computing capabilities. In the system architecture of autonomous driving solutions, tasks are generally divided into multiple modules, and a series of sensors and algorithms are used on each module. The use of multiple modules can realize control and constraints for autonomous vehicles at different angles, reducing the error rate of autonomous driving and increasing system security. Autonomous driving in special scenarios and complex environments has always been a ...

AI Technical Summary

Problems solved by technology

[0003] The present invention provides an automatic driving auxiliary safety early warning system in dark scenes and its early warning method. The early warning system is aimed at dark scenes, such as dark nights, dark parking lots with blind spots, and a series of scenes, which can realize a detection area. Self-defined semantic segmentation with a low amount of calculation can effectively realize automatic car-following, judgment of the driving intention of the vehicle ahead and anti-collision warning, pedestrian target detection and anti-collision warning, estimation of vehicle speed and orientation in the lateral blind zone and anti-collision warning At the same time, it can make up for the low sensitivity of visible light cameras in dark scenes when the light intensity changes sharply. In dark scenes, it can detect unknown vehicles coming from the side blind area, issue safety warnings, and reduce the traffic accident rate.

Images