Self-cleaning method and system for view field area of front-view camera of windshield

Abstract

The invention discloses a self-cleaning method and system for a view field area of a front-view camera of a windshield. The method comprises the steps that: image information in the view field area of the front-view camera of the windshield of a vehicle is collected; according to the image information, whether polluted foreign matter exists in the view field area of the front-view camera of the windshield of the vehicle or not is judged; and if the polluted foreign matter exists in the view field area of the front-view camera of the windshield of the vehicle, the polluted foreign matter is cleaned through a windscreen wiper system and a washing system. According to the self-cleaning method and system for the view field area of the front-view camera of the windshield, when it is judged that the polluted foreign matter exists in the view field area of the front-view camera of the vehicle windshield according to the image information, cleaning is conducted through the windscreen wiper system and the washing system, it can be guaranteed that the windshield area of the front-view camera is clean, the image acquisition quality of the foresight camera is improved, and normal work of an advanced driving assistance system is guaranteed; and automatic cleaning of the corresponding polluted area is achieved, omission and inconvenience caused by manual operation are avoided, and the intelligent vehicle using experience of a user is improved.

Description

technical field [0001] The invention relates to the technical field of vehicle self-cleaning, in particular to a self-cleaning method and system for the field of view of a windshield front-view camera. Background technique [0002] With the rapid development of artificial intelligence and image processing technology applications in recent years, the automobile industry has ushered in a new revolution in intelligence. The perception of road traffic information through sensors such as cameras and radars is the basic link of intelligent driving. Among them, the vehicle-mounted front-view camera plays an important role in lane line recognition and vehicle pedestrian recognition in front. Monocular cameras need to collect a large number of samples for image recognition and classification, and generally estimate the distance based on the size of the target image, which has a large uncertainty; in contrast, binocular cameras directly use the parallax of the two cameras to calculat...

AI Technical Summary

Problems solved by technology

[0003] At this stage, most OEMs choose to place the front-view camera on the inside of the front windshield. The windshield within the field of view of the front-view camera is prone to stains and rain, which will adversely affect the quality of the collected images and reduce the road traffic ahead. Accuracy of situation recognition, deepening the safety hazard of vehicle driving

Images