Direct Triangulation with Spherical Projection for Omnidirectional Cameras
This provides a more efficient triangulation method for computer vision applications using omnidirectional cameras, though it appears incremental as it builds on existing optimization approaches.
The paper tackles the problem of triangulation for calibrated omnidirectional cameras by optimizing ray-pairs on the projective sphere, resulting in a closed-form, non-iterative solution that is computationally inexpensive compared to previous methods. It also demonstrates applicability to non-omnidirectional cameras.
In this paper, it is proposed to solve the problem of triangulation for calibrated omnidirectional cameras through the optimisation of ray-pairs on the projective sphere. The proposed solution boils down to finding the roots of a quadratic function, and as such is closed form, completely non-iterative and computationally inexpensive when compared to previous methods. In addition, even thought the motivation is clearly to solve the triangulation problem for omnidirectional cameras, it is demonstrated that the proposed methods can be applied to non-omnidirectional, narrow field-of-view cameras.