Automatic detection and correction method of external parameters of vehicle-mounted camera

Abstract

The invention discloses a method for automatic detection and correction of external parameters of a vehicle-mounted camera, and the method comprises the steps: S1, capturing a real scene image after a vehicle-mounted camera application is started, extracting feature points of the real scene image, calling the information of a vehicle body fixed part feature point set of the initial image in a storage unit, and carrying out the feature point matching between the extracted feature points in the real scene image and the information of the vehicle body fixed part feature point set of the initial image, wherein the real scene image comprises an image of a vehicle body fixed part in an initial state; S2, calculating the mean error E between the matched feature point sets: executing step S3 if the mean error E is greater than or equal to an allowed error bound e, or else executing step S4; S3, calculating an external parameter error compensation matrix of the vehicle-mounted camera, correcting the external parameters of the vehicle-mounted camera, and obtaining a corrected camera model; S4, operating a camera application system. According to the invention, the method enables the vehicle-mounted camera to remain in a good working state, and reduces the errors caused by the changes of the position or angle of the camera.

Description

technical field [0001] The invention belongs to the technical field of computer vision measurement. The specific content is an automatic detection and correction method for the external parameters of the camera proposed for the position offset problem caused by external reasons after the calibration of the intelligent vehicle vision sensor (camera) parameters. Background technique [0002] In the high-tech fields such as intelligent vehicle automatic driving technology and industrial robot control technology, visual measurement and environmental perception technology are the key to realizing the technical basis and intelligent characteristics of intelligent system perception, cognition, decision-making and control. Among them, visual sensors (industrial cameras, Compared with acoustic sensors, millimeter-wave radar, and laser radar, etc., it has the advantages of large amount of information, low price, wide detection range, and strong anti-interference ability, and has been ...

AI Technical Summary

Problems solved by technology

However, during the long-term application of the vehicle, it is easy to happen due to road bumps and other reasons, especially smart cars need to use the on-board camera for more accurate target recognition, environment perception and vision-based vehicle high-precision Positioning, this kind of offset camera must be re-calibrated with external parameters

However, under normal circumstances, users do not have the technical conditions to perform parameter calibration according to the method and process before the installation and application of the vehicle vision sensor system. It is very necessary to research and propose the technology of relative position deflection detection and external parameter automatic correction during the use of vehicle cameras. , which has obvious engineering practical value for ensuring the accuracy of vehicle-mounted camera vision measurement and environmental perception, and improving the robust performance of the system

[0004] According to domestic and foreign scientific and technological literature and public patent information retrieval, no one has proposed the installation and use of visual sensors (such as vehicle cameras, etc.), without the aid of any calibration tools (such as calibration plates, flat or spherical mirrors, etc.), no need Under the condition of information such as the position coordinates of the reference object, the method of automatically detecting and correcting the external parameters of the camera

Images