Spherical Formulation of Geometric Motion Segmentation Constraints in Fisheye Cameras

We introduce a visual motion segmentation method employing spherical geometry for fisheye cameras and automated driving. Three commonly used geometric constraints in pin-hole imagery (the positive height, depth epipolar constraints) are reformulated to coordinates, making them invariant specific camera configurations as long the calibration is known. A fourth constraint, known anti-parallel added resolve motion-parallax ambiguity, support detection of moving objects undergoing parallel or near-parallel with respect host vehicle. final constraint described, three-view described though not employed our proposed algorithm. Results presented analyzed that demonstrate proposal an effective approach direct employment on imagery.