An analytical technique for boundary reconstruction in occluded objects

A real-valued function which is defined in some closed interval is considered to model the boundary of a physical object. Given such a model, it is assumed that in some subset of the domain the values of the function are not known. This case is considered to represent the physical situation where a section of the object's boundary is occluded or hidden. The authors present a scheme for reconstructing the occluded boundary of such an object, that is, for identifying the unknown values of the function. The scheme utilizes some analytical properties of the known function values to define a range of possible values for the unknown function values. The scheme is simple, efficient and, in view of the limitations presented by these two characteristics, apparently performs reasonably well.<>