Today I read a paper titled “Differential Methods in Catadioptric Sensor Design with Applications to Panoramic Imaging”
The abstract is:
We discuss design techniques for catadioptric sensors that realize given projections.
In general, these problems do not have solutions, but approximate solutions may often be found that are visually acceptable.
There are several methods to approach this problem, but here we focus on what we call the “vector field approach”.
An application is given where a true panoramic mirror is derived, i.e.
a mirror that yields a cylindrical projection to the viewer without any digital unwarping.