Calibration of parabolic refraction and reflection camera through Veronese mapping and checkerboard

Abstract

The invention relates to a method for calibrating a parabolic refraction and reflection camera through the Veronese mapping and a checkerboard. The method is characterized in that the target of the method is formed by a known checkerboard in the space. The method comprises the specific steps of during the solving process of the internal parameters of a parabolic refraction and reflection camera, shooting 3 images by the parabolic refraction and reflection camera; obtaining three homography matrixes through the Veronese mapping so as to solve a vanishing point for any straight line; solving another vanishing point based on the relationship between antipodal points and harmonic conjugates, wherein the connecting line of the two vanishing points is a vanishing line; according to the intersection points of the vanishing line with the image of the straight line namely a quadratic curve, solving the image of three sets of circular points; based on the constraint of the image of circular points on the image of the absolute quadratic curve, solving the internal parameters of the parabolic refraction and reflection camera through the Cholesky decomposition.

Description

technical field [0001] The invention belongs to the field of computer vision, and relates to a homography matrix from a plane grid to an image plane in a parabolic catadioptric camera by using Veronese mapping, and calibrates the catadioptric camera by combining vanishing points and ring points. Background technique [0002] The research goal of computer vision is to enable computers to have the ability to recognize three-dimensional environmental information through two-dimensional images. This ability will not only enable the machine to perceive the geometric information (shape, position, posture, motion, etc.) of objects in a three-dimensional environment, but also describe, store, recognize and understand them. In this process, the mapping process between the three-dimensional calibration object and its two-dimensional image must be determined. In order to determine this mapping process, the geometric imaging model of the camera needs to be established. The parameters of...

AI Technical Summary

Problems solved by technology

The literature "Plane-based calibration of central catadioptric cameras", (GaspariniS., SturmP., BarretoJ.P., IEEE12thInternationalConferenceon, pp.1195-1202, 2009) first passes at least three single The image of the absolute conic curve (IAC) is obtained by calculating the matrix, and then the internal parameters can be obtained according to the relationship between IAC and the internal parameters of the central catadioptric camera, but this method requires iterative estimation

Images