Abstract:Central catadioptric cameras are imaging devices that use mirrors to enhance the field of view while preserving a single effective viewpoint. Lines and spheres in space are all projected into conics in the central catadioptric image planes, and such conics are called line images and sphere images, respectively. We discovered that there exists an imaginary conic in the central catadioptric image planes, defined as the modified image of the absolute conic (MIAC), and by utilizing the MIAC, the novel identical projective geometric properties of line images and sphere images may be exploited: Each line image or each sphere image is double-contact with the MIAC, which is an analogy of the discovery in pinhole camera that the image of the absolute conic (IAC) is double-contact with sphere images. Note that the IAC also exists in the central catadioptric image plane, but it does not have the double-contact properties with line images or sphere images. This is the main reason to propose the MIAC. From these geometric properties with the MIAC, two linear calibration methods for central catadioptric cameras using sphere images as well as using line images are proposed in the same framework. Note that there are many linear approaches to central catadioptric camera calibration using line images. It seems that to use the properties that line images are tangent to the MIAC only leads to an alternative geometric construction for calibration. However, for sphere images, there are only some nonlinear calibration methods in literature. Therefore, to propose linear methods for sphere images may be the main contribution of this paper. Our new algorithms have been tested in extensive experiments with respect to noise sensitivity.

Metric Planar Rectification from Perspective View Via Circles.

Using Coplanar Circles to Perform Calibration-Free Planar Scene Analysis under a Perspective View

Efficient Extraction of Metric Measurements for Planar Scene under 2D Homography with the Help of Planar Circles

A Simple 3D Reconstruction Method Based on Single Camera

Camera lens radial distortion correction using two-view projective invariants.

Planar Rectification by Solving the Intersection of Two Circles under 2D Homography.

Planar Metric Rectification by Algebraically Estimating the Image of the Absolute Conic

Centra Catadioptric Camera Calibration with Single Image.

Unwrapping and Stereo Rectification for Omnidirectional Images

Planar Visual Metrology Using Partition-Based Camera Calibration

Sketch-based 3-D modeling for piecewise planar objects in single images

3D Information Acquisition from Rectangles and Circles in Single Images

Fast Projective Image Rectification for Planar Objects with Manhattan Structure

Efficient Detection of Projected Concentric Circles Using Four Intersection Points on a Secant Line

Stereo Image Rectification Algorithm for Multi-View 3D Display

Full Camera Calibration from a Single View of Planar Scene

Perspective Distortion Rectification For Planar Object Based On Lidar And Camera Data Fusion

Single View Metrology along Orthogonal Directions

Identical Projective Geometric Properties of Central Catadioptric Line Images and Sphere Images with Applications to Calibration

A rectification algorithm for un-calibrated multi-view images based on SIFT features

Distance Measurement from Single Image Based on Circles