Image Reconstruction from Gabor Magnitudes
We present an analysis of the representation of images as the magnitudes of their transform with complex-valued Gabor wavelets. Such a representation is very useful for image understanding purposes and serves as a model for an early stage of human visual processing. We show that if the sampling of the wavelet transform is appropriate then the reconstruction from the magnitudes is unique up to the sign for almost all images. We also present an iterative reconstruction algorithm derived from the ideas of the proof, which yields very good reconstruction up to the sign and minor numerical errors in the very low frequencies.
KeywordsGabor wavelets reconstruction feature extraction phase retrieval Gabor magnitudes visual cortex
Unable to display preview. Download preview PDF.
- 6.Monson H. Hayes and James H. McClellan. Reducible Polynomials in More Than One Variable. Proceedings of the IEEE, 70(2):197–198, February 1982.Google Scholar
- 7.J.P. Jones and L.A. Palmer. An evaluation of the two-dimensional Gabor filter model of simple receptive fields in cat striate cortex. Journal of Neurophysiology, 58(6):1233–1258, 1987.Google Scholar
- 8.Gerald Kaiser. A Friendly Guide to Wavelets. Birkhäuser, 1994.Google Scholar
- 10.R. Murenzi. Wavelet Transforms Associated to the n-Dimensional Euclidean Group with Dilations: Signal in More Than One Dimension. In J. M. Combes, A. Grossmann, and P. Tchamitchian, editors, Wavelets—Time-Frequency Methods and Phase Space, pages 239–246. Springer, 1989.Google Scholar
- 11.Alan V. Oppenheim and Jae S. Lim. The Importance of Phase in Signals. Proceedings of the IEEE, 96(5):529–541, May 1981.Google Scholar
- 14.Jochen Triesch and Christoph von der Malsburg. Robust classification of hand postures against complex backgrounds. In Proceedings of the Second International Conference on Automatic Face and Gesture Recognition, pages 170–175. IEEE Computer Society Press, 1996.Google Scholar
- 16.Christoph von der Malsburg and Ladan Shams. Role of complex cells in object recognition. Nature Neuroscience, 2001. Submitted.Google Scholar
- 19.Ingo J. Wundrich, Christoph von der Malsburg, and Rolf P. Würtz. Image representation by the magnitude of the discrete Gabor wavelet transform. IEEE Transactions on Image Processing, 1999. In revision.Google Scholar