Geometric Interpretations of the Normalized Epipolar Error
This provides theoretical insights for researchers in computer vision, particularly in 3D reconstruction and multi-view geometry, but is incremental as it builds on existing error metrics.
The paper tackles the problem of interpreting the normalized epipolar error in computer vision by showing it relates to geometric quantities such as the shortest distance between backprojected rays, the dihedral angle between epipolar planes, and the L1-optimal angular reprojection error.
In this work, we provide geometric interpretations of the normalized epipolar error. Most notably, we show that it is directly related to the following quantities: (1) the shortest distance between the two backprojected rays, (2) the dihedral angle between the two bounding epipolar planes, and (3) the $L_1$-optimal angular reprojection error.