Matching of stereo image pairs for industrial vision systems

A matching technique for stereo images which uses the Fourier transform for disparity estimation in a robot workspace environment is presented. The method consists of three phases. The first is a disparity estimation process which makes use of the shifting property of the Fourier transform. The second is a guided feature matching technique which selects the corners as feature points. The third is a local matching procedure that uses the corner matching results. The final output of the method is a disparity map corresponding to the matched points. The results show that the method produces accurate and dense disparity maps in a robust workspace environment.