Camera distortion correction method of distortion-free model

Abstract

The invention discloses a camera distortion correction method for a distortion-free model, and the method comprises the steps: firstly enabling a computer to generate a longitudinal standard stripe image and a transverse standard stripe image as correction templates, displaying the correction templates on a flat-panel display, enabling a camera to shoot the images, and obtaining a distorted image;respectively carrying out phase shift phase solution and phase unwrapping on the two groups of distorted images to obtain phase distribution of the distorted images; obtaining the instantaneous spatial frequency distribution of the distorted image, and determining the instantaneous frequency of a distortion center stripe; extracting distortion image phase values of continuous multiple points nearthe distortion center, and performing linear fitting to obtain a phase value of the distortionless image; subtracting the phase distribution of the distorted image from the phase distribution of thedistortionless image to obtain distorted phase distribution of the distorted image, and then obtaining distortion amount distribution of the distorted image; acquiring distortion distribution under acorrection coordinate system and a correction mapping graph; and finally, carrying out distortion correction on the image shot by the camera by using the correction mapping graph. According to the invention, the distorted camera image can be corrected without a distortion model.

Description

technical field [0001] The invention relates to the technical field of computer vision distortion image correction and optical measurement, in particular to a camera distortion correction method without distortion model. Background technique [0002] Computer vision is to capture image information through various camera lenses and equipment, analyze and process it, and imitate human beings to realize image recognition and scene reconstruction. In order to meet the capture of scene information with a large viewing angle, lenses with wide field of view imaging characteristics are widely used in various fields. However, there are various degrees of distortion in images captured by wide-field lenses, the most notable of which is radial distortion. Radial distortion is a kind of position offset along the radial direction of the distortion center. The distortion makes it impossible to accurately determine the geometric position and size of the object in the image, and even causes...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to propose a camera distortion correction method that does not depend on the distortion model, so as to overcome the defects of current camera distortion correction using the distortion model

Images