Vehicle-mounted monocular camera external parameter self-calibration method

Abstract

The present invention discloses a vehicle-mounted monocular camera external parameter self-calibration method. The method comprises the following steps that: 1) the image sequence of a road surface on which an automobile travels is shot through a vehicle-mounted camera, wherein the vehicle-mounted camera is installed in the central position of the front windshield of the automobile; 2) based on the acquired image sequence, an image segmentation algorithm is used to extract lane mark feature points, two parallel lane lines are obtained according to the fitting of the feature points, a road mark model is established, and vanishing points are determined; 3) the motion vectors of the lane mark feature points are matched through a feature point matching method, and points of which the longitudinal motion vectors are zero are extracted, and the extracted points are subjected to fitting, so that a motion direction line can be obtained, and the yaw angle of the camera is calculated; 4) the extension focal length of an image is determined; and 5) the external parameters of the camera are calculated according to the extension focal length and a known image main point. According to the vehicle-mounted monocular camera external parameter self-calibration method of the invention, man-made manipulation procedures are not required; the implementation of the vehicle-mounted monocular camera external parameter self-calibration method is convenient and quick, it only needs to make the automobile travel for a certain time, and the parameter calibration result of the camera can be obtained through calculation.

Description

technical field [0001] The invention relates to automobile driving assistance technology, in particular to a method for self-calibrating external parameters of a vehicle-mounted monocular camera. Background technique [0002] The traditional camera calibration method is a relatively strict calibration method, so its accuracy is high, but the calibration process is more complicated, and the amount of calculation is large, so it has limitations in the use of vehicle cameras that are widely used. [0003] For the existing self-calibration methods, it is generally not easy to find a scene with three parallel line pairs in the vertical direction. Even if such a scene condition exists, due to the influence of environmental factors, there will be problems in the extraction accuracy of the straight line edge, which will affect the The accuracy of camera calibration parameter calibration. [0004] The present invention does not completely depend on specific scene markers, and uses m...

AI Technical Summary

Problems solved by technology

[0002] The traditional camera calibration method is a relatively strict calibration method, so its accuracy is high, but the calibration process is more complicated, and the amount of calculation is large, so it has limitations in the use of widely used vehicle cameras

[0003] For the existing self-calibration methods, it is generally not easy to find a scene with three parallel line pairs in the vertical direction. Even if such a scene condition exists, due to the influence of environmental factors, there will be problems in the extraction accuracy of the straight line edge, which will affect the Calibration accuracy of camera calibration parameters

Images