Channel capacity model of binary encoded structured light-stripe illumination.

A common approach to structured light-illumination measurement is to encode a surface topology successively with binary light-stripe patterns of variable spatial frequency. Each surface location is thereby encoded with a binary sequence associated with its height. By analyzing the lateral displacements of the reflected encoded pattern, one can reconstruct the surface topology without ambiguity. We present a model for multistripe analysis in terms of an information channel for which the maximum spatial stripe frequency is related to channel capacity and maximized accordingly by use of Shannon's theorems. The objective is to improve lateral resolution through optimized spatial frequency while maintaining a fixed range resolution. Given an optimized spatial frequency, a technique is presented to enhance lateral resolution further by multiplexing the light structure. Theoretical and numerical results are compared with experimental data.