Method and device for calculating car distance and collision time

Abstract

The invention relates to a method and device for calculating car distance and collision time with functions of improving car distance calculation accuracy and performing accurate real-time early warning. The method comprises the steps of image acquisition, wherein a vehicle-mounted camera of the own vehicle acquires an image of a front car; front car detection and tracking, wherein the image is processed, and position information of feature points of the front car is detected; front car distance calculation, wherein the position information of the feature points of the front car in a predetermined group is taken, the width of the front car and the car distance are calculated to obtain an average width value of the front car, and the car distance between the own car and the front car is corrected according to the average width value of the front car; front car collision early warning, wherein a relative car distance curve is acquired, real-time relative speed is obtained according to the car distance curve, collision time is calculated, the collision time is compared with an early warning threshold, and it is judged whether or not early warning is triggered; the relative car distance curve is obtained by fitting a Bezier spline, and the corrected car distance is taken as an input value of the Bezier spline.

Description

technical field [0001] The present application relates to the technical field of intelligent transportation, in particular to a method and device for calculating vehicle distance and collision time. Background technique [0002] Existing vehicle-mounted systems generally use vehicle-mounted radar to measure the distance of vehicles in front, or use ultrasonic sensors and radio frequency transmitters to detect the distance of front and rear obstacles, and calculate the collision time to issue collision warnings. However, such equipment is expensive and not easy to promote. [0003] There is also a vehicle ranging and collision warning based on monocular vision: a single camera collects the front image and runs the training sample method to detect the front target. The distance between the two vehicles is calculated according to the principle of small hole imaging and the geometric ratio, and the collision time is calculated in combination with the current speed of the vehicle...

AI Technical Summary

Problems solved by technology

However, the distance calculated entirely by the vehicle detection position in the image, the small hole imaging principle and the geometric ratio relationship will have a large error in practical application due to the uneven road surface or the error of vehicle detection. The error calculated by the small hole imaging principle It is proportional to the distance, that is, the closer it is, the smaller it is, and the farther it is, it is larger

Images