MERGING 3D POINT CLOUDS WITH RADIOMETRIC ENHANCEMENT

Modules for radiometric enhancement of colored 3D point clouds by propagating colors

This implementation includes modules for radiometric enhancement of colored 3D point clouds and nearest neighbor search. The introduced method detects the overlapping parts and derives a transformation function that reduces radiometric differences. This transformation function can be in one of the following forms: linear contrast stretching or histogram matching. Overlap detection is performed as radial search procedure, in which the first closest neighbor is looked for. Available search strategies are brute-force search, hash grid search, search using octrees and kdtrees. See other postings of the author for very fast loading and saving of colored and bulky point clouds. The implementation is based on "Merging 3D Point Clouds with Radiometric Enhancement", U. Yilmaz and O. Hellwich, Proceedings of CGV-CON, 2010.
USAGE
-----
cnnsearch --------> searches the nearest neighbors of a point cloud in another (Function definitions in cnnsearch.cpp)
cnnsearchTest ----> is an auxiliary module to show how cnnsearch works.
cpropagate -------> enhances point clouds radiometrically by propagating.
cpropagateTest ---> is an auxiliary module to show how cpropagate works.