A lane departure warning method based on vehicle-mounted binocular camera

Abstract

The invention relates to a vehicle-mounted binocular camera based lane departure early-warning method. The vehicle-mounted binocular camera based lane departure early-warning method comprises the steps that a vehicle passable region is estimated based on a vehicle-mounted binocular camera; lane line detection and tracking are performed in the estimated vehicle passable region; a lane departure distance D is calculated; a vehicle side departure angle beta is calculated; improved lane departure time (TLC) is calculated; the lane departure distance and the improved lane departure time are comprehensively considered, and early vision warning and early hearing warning combined lane departure graded early warning signals are correspondingly sent to remind a driver to make corresponding measures in time. The vehicle-mounted binocular camera based lane departure early-warning method is high in detection accurate rate and greatly improves the accuracy and reliability of lane departure early-warning, the early warning signals have gradient, no sudden occurs, the driver response degree is higher, corresponding measures can be accurately adopted, and accordingly the vehicle driving safety is improved.

Description

technical field [0001] The invention relates to the field of vehicle control, in particular to a lane departure warning method based on a vehicle-mounted binocular camera. Background technique [0002] Lane departure warning is one of the main assisted driving functions of the vehicle. It warns the lane departure phenomenon when the driver has no intention to change lanes, so as to avoid potential lane change traffic accidents or illegal driving behaviors (solid line lane change). The existing lane departure warning system is mainly based on the image analysis and processing of the monocular camera, extracting lane line information, calculating warning indicators and making warning decisions. However, the existing lane departure warning has the following disadvantages: lane line detection is easily affected by light, shadow, etc.; based on assumptions such as plane and straight line markings, the application range is limited; the lane line detection range is a fixed image ar...

AI Technical Summary

Problems solved by technology

However, the existing lane departure warning has the following disadvantages: lane line detection is easily affected by light, shadow, etc.; based on assumptions such as plane and straight line markings, the application range is limited; the lane line detection range is a fixed image area, and the detection results are easily affected by this area. The impact of other vehicles appearing within; the early warning reference index is single, and the driver's manipulation correction is not considered, and the false alarm rate is high; the motion estimation does not consider the vehicle steering operation; a single-level early warning signal is adopted, and the early warning signal has no gradient gradient, and the occurrence is relatively sudden. Poor comfort, poor driver feedback

Images