The translation sensitivity of wavelet-based registration

in IEEE Transactions on Pattern Analysis and Machine Intelligence

Harold S. Stone, NEC
Jacqueline Le Moigne
Morgan McGuire, MIT (currently at Williams College)

Paper (PDF) Abstract


This paper studies the effects of image translation on wavelet-based image registration. The main result is that the normalized correlation coefficients of low-pass Haar and Daubechies wavelet subbands are essentially insensitive to translations for features larger than twice the wavelet blocksize. The third-level low-pass subbands produce a correlation peak that varies with translation from 0.7 and 1.0 with an average in excess of 0.9. Translation sensitivity is limited to the high-pass subband and even this subband is potentially useful. The correlation peak for high-pass subbands derived from first and second-level low-pass subbands ranges from about 0.0 to 1.0 with an average of about 0.5 for Daubechies and 0.7 for Haar. We use a mathematical model to develop these results, and confirm them on real data.

This is an substantially expanded and revised version of: Stone, LeMoigne, and McGuire. The translation sensitivity of wavelet-based registration. Proceedings of the 26th AIPR Workshop, Proceedings of the SPIE, Exploiting New Image Sources and Sensors 3240:116-125, 1997


 author = {Harold S. Stone and Jacqueline Le Moigne and Morgan McGuire},
 title = {The Translation Sensitivity of Wavelet-Based Registration},
 journal = {IEEE Trans. Pattern Anal. Mach. Intell.},
 volume = {21},
 number = {10},
 year = {1999},
 issn = {0162-8828},
 pages = {1074--1081},
 doi = {},
 publisher = {IEEE Computer Society},
 address = {Washington, DC, USA},