EHSAN SADEGHI POOR - Computer and Electronic Department , Hormozgan University , Bandar abbas , IRAN
NASROLAH SAHRAGARD - Computer and Electronic Department , Hormozgan University , Bandar abbas , IRAN

This paper presents a novel algorithm of stereo correspondence with rank segmentation. In this algorithm we used the genetic algorithm to achieve the accurate disparity map. This method produce the accurate disparity map compared with other methods and this is very important to compose the 3D model. Genetic algorithms are efficient search methods based on principles of population genetic, i.e. mating, chromosome crossover, gene mutation, and natural selection. Finally morphology is employed to remove the errors and discontinuities. The applied method generated good simulation results compared to those obtained by correlation using fixed sized window.

genetic algorithm, morphology, segmentation, stereo correspondence