Binocular distance measurement method for vehicle blind area detection alarm device

Abstract

The invention provides a binocular distance measurement method for a vehicle blind area detection alarm device. The method comprises the following steps: calibrating the parameters of a binocular camera; measuring the specific value of the rotation angle of the binocular camera relative to the vehicle body side; inputting an image acquired by the binocular camera into a target detection algorithm,outputting a target by the target detection algorithm, and determining position information of the target in the image; calling a binocular stereo matching algorithm to determine parallax of the target in the two frames of images; determining a first distance from a projection point of the target on a plane where a connecting line between the optical axis and the lens is located to a vertical point of the connecting line between the lenses based on the focal length, the parallax and the distance between the lenses; obtaining a vertical distance from the target to the vehicle body based on therotation angle, the rolling angle, the coordinate of the target in the image coordinate system, the coordinate of the optical center, the focal length and the target first distance; the vertical distance between the target and the vehicle is the shortest distance between the target and the vehicle, and the occurrence probability of traffic accidents can be effectively reduced when the method is applied to a blind area early-warning system.

Description

technical field [0001] The invention relates to the technical field of auxiliary equipment for safe driving of automobiles, in particular to a binocular ranging method for a vehicle blind spot detection and alarm device. Background technique [0002] The blind spot of the vehicle is the blind spot of sight, usually in the rearview mirror or the A-pillar, etc. The existence of the blind spot makes it impossible for the driver to understand the situation in the blind spot. Once an accident occurs, traffic accidents are very likely to occur. Designed for visual blind spots, [0003] One type of existing vehicle blind spot detection system uses sensors for detection, and the other type uses multiple monocular cameras for detection. Sensor detection cannot distinguish people or objects, and is prone to false detection. It can only detect the distance from the sensor to the target. However, the vertical distance from the vehicle to the target cannot be obtained, and it is easy to...

AI Technical Summary

Problems solved by technology

[0003] One type of existing vehicle blind spot detection system uses sensors for detection, and the other type uses multiple monocular cameras for detection. Sensor detection cannot distinguish people or objects, and false detection is prone to occur. It can only detect the distance from the sensor to the target. However, the vertical distance from the vehicle to the target cannot be obtained, and large errors are likely to occur. If a good detection accuracy is to be obtained, a large number of equipment needs to be installed and the cost is high;

[0004] Another type of solution requires manual pre-calibration of the detection area given by the blind spot. However, this solution still cannot give accurate distance information. It can only give the straight-line distance from the camera to the target, and get the plane where the line connecting the object to the lens is located. distance, not the vertical distance to the vehicle

[0005] Because the vertical distance from the target to the vehicle is the shortest distance between the target and the vehicle, how to obtain the shortest distance is very important for vehicle blind spot detection, and the detection accuracy of the blind spot warning system plays a significant role in avoiding traffic accidents. However, the existing vehicle blind spot detection system Neither can detect the vertical distance from the target to the vehicle

Images