Effects of the change of window size on the performance of correlation-based stereo

One of the difficulties in the correlation-based stereo matching is the selection of window size, since there are two competing factors - perspective distortion and stereo matching error. This paper suggests a tool to tune the correlation window size in a given image domain by using photo-realistic simulation methodology. To facilitate the analysis proposed in this paper, we use photo-realistic simulation methodology to generate a pair of photo-realistic synthetic images of the terrain from a pre-acquired DEM(Digital Elevation Map) and ortho-image, which can be served as the pseudo ground truth. We performed 3D reconstruction on synthetic images of a natural terrain and carried out the evaluation of the correlation window on DEM accuracy. Experimental results are consistent with our strong expectation about two competing factors and show that our approach can be a useful tool to tune the window size.