Stereo-Matching Techniques Optimisation Using Evolutionary Algorithms

摘要

In this paper we present a novel approach to 3D stereo-matching which uses an evolutionary algorithm in order to optimise 3D reconstruction. Common techniques in the field of 3D models generation are employed together with a Genetic Algorithm (GA) which is able to improve the results of the matching process. A general overview of the most relevant approaches is given in order to contextualise our method and to analyse its strength-points and potentialities. Details of the implemented GA are discussed with a particular focus on the constraints used in order to obtain better results. Experimental results of the trials carried out are given in a final stage together with concluding remarks and some cues for further research.