Abstract: This work introduces a solution based on genetic algorithms to find the overlapping area between two point cloud captures obtained from a three-dimensional scanner. Considering three translation coordinates and three rotation angles, the genetic algorithm evaluates the matching points in the overlapping area between the two captures given that transformation. Genetic simulated annealing is used to improve the accuracy of the results obtained by the genetic algorithm.

Keywords: Range images, genetic algorithms, overlaping point cloud captures, genetic simulated annealing.

Link:

OVERLAPPING RANGE IMAGES USING GENETIC ALGORITHMS

Fernando Ortega, Javier San Juan, Francisco Serradilla, Dionisio Cortes

http://www.foibg.com/ijita/vol20/ijita20-03-p06.pdf