Optimization of stereo disparity estimation using the instantaneous frequency
The use of phase differences from Gabor filter responses is a well established technique for the computation of stereo disparity. It achieves the subpixel estimation of disparity without applying a correspondence search. However, the problem of scale or central frequency selection is still unsolved. Here, we study the effects of varying filter frequency on the disparity estimation and we compare it to the use of the instantaneous frequency. The analytical results on several models of intensity and disparity variation show robustness of the disparity estimates against variations in the filter-wavelength.
KeywordsInstantaneous Frequency Gabor Filter Constant Model Local Phase Signal Wavelength
Unable to display preview. Download preview PDF.
- 4.K. Langley, T.J. Atherton, R.G. Wilson, and M.H.E. Larcombe. Vertical and horizontal disparities from phase. In Proc. First European Conference on Computer Vision, pages 315–325. Antibes, France, Apr. 23–26, O.D. Faugeras (Ed.), Lecture Notes in Computer Science 427, Springer-Verlag, Berlin et al., 1990.Google Scholar
- 5.A. Maki, L. Bretzner, and J.-O. Eklundh. Local fourier phase and disparity estimation: An analytical study. In V. Hlavac et al. (Ed.), Pro c. Int. Conf. Computer Analysis of Images and Patterns CAIP, Prag, pages 868–874, 1995.Google Scholar
- 7.W. M. Theimer and H. P. Mallot. Phase-based binocular vergence control and depth reconstruction using active vision. CVGIP: Image Understanding, 60(2):343–358, 12 1994.Google Scholar
- 8.C.J. Westelius, H. Knutsson, J. Wiklund, and C.F. Westin. Phased-based disparity estimation. In J.L. Crowley and H.I. Christensen, editors, Vision as Process, pages 157–178. Springer Verlag, Berlin et al., 1994.Google Scholar