Share Email Print
cover

Proceedings Paper

Power cepstral image analysis through the scale transform
Author(s): Jose Luis Pech-Pacheco; Gabriel Cristobal; Josue Alvarez-Borrego; Leon Cohen
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The detection of image changes irrespective on geometric transformations are required in many applications. In this paper we present a novel use of the scale transform oriented to image identification and registration. If we translate a signal then all the information appears in the phase of the Fourier transform of the translated signal. Similarly, if we scale or rotate an image all the information about the amount of scaling or rotation appear in the phase of the scale transform. In the present study we report a very precise image identification technique based on the use of the power cepstrum of the scale transform. Cepstral filtering can be considered as a non-linear adaptive prefilter followed by an autocorrelation operation. The accuracy of the cepstrum techniques and the speed of the Fourier transform makes the present method faster and more robust to noise than other existing techniques. Image registration has been accomplished by computing the power cepstrum of the log-polar scale mapping. The performance of the improved method has been experimentally verified in a class of typed characters and diatom images in lighting microscopy.

Paper Details

Date Published: 2 November 2000
PDF: 12 pages
Proc. SPIE 4113, Algorithms and Systems for Optical Information Processing IV, (2 November 2000); doi: 10.1117/12.405864
Show Author Affiliations
Jose Luis Pech-Pacheco, Instituto de Optica (Spain)
Gabriel Cristobal, Instituto de Optica (Spain)
Josue Alvarez-Borrego, CICESE (Mexico)
Leon Cohen, CUNY/Hunter College and Graduate Ctr, (United States)


Published in SPIE Proceedings Vol. 4113:
Algorithms and Systems for Optical Information Processing IV
Bahram Javidi; Demetri Psaltis, Editor(s)

© SPIE. Terms of Use
Back to Top