Shape from Shading using Parallel Multigrid Relaxation
Wim Sweldens, Dirk Roose · Birkhäuser Basel eBooks · 1991
The aim of a shape from shading algorithm is to reconstruct the shape of an object from a two-dimensional image of that object. This can be formulated as a variational problem that leads to a system of non-linear partial differential equations. In order to find a solution, accurate information about the occluding boundary was required in early methods. This is often not available or difficult to obtain. Omitting occluding boundary information for scenes lighted from overhead leads to ambiguous problems that cannot be solved with standard relaxation methods. This difficulty can be overcome by using the full multigrid scheme. Since most of the computations are local, grid partitioning and parallel processing can be exploited successfully. We report on results for synthetic and camera images, obtained on the Intel iPSC/2, a distributed memory parallel computer. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.