The Translation Sensitivity of Wavelet-Based Registration

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.

[1]  Lisa M. Brown,et al.  A survey of image registration techniques , 1992, CSUR.

[2]  Rama Chellappa,et al.  A computational vision approach to image registration , 1993, IEEE Trans. Image Process..

[3]  Arthur P. Cracknell,et al.  Pixel and sub-pixel accuracy in geometrical correction of AVHRR imagery , 1989 .

[4]  P. E. Anuta,et al.  Spatial Registration of Multispectral and Multitemporal Digital Imagery Using Fast Fourier Transform Techniques , 1970 .

[5]  Harold S. Stone,et al.  Progressive wavelet correlation using Fourier methods , 1999, IEEE Trans. Signal Process..

[6]  Gian Antonio Mian,et al.  On the registration of an object translating on a static background , 1996, Pattern Recognit..

[7]  Rajesh Shenoy,et al.  New Gabor wavelets with shift-invariance for improved time-frequency analysis and signal detection , 1996, Defense + Commercial Sensing.

[8]  Paul D. Fiore.  Image registration using both distance and angle information , 1995, Proceedings., International Conference on Image Processing.

[9]  Truong Q. Nguyen,et al.  Wavelets and filter banks , 1996 .

[10]  Yi-Tong Zhou,et al.  Wavelet-based point feature extractor for multisensor image restoration , 1996, Defense + Commercial Sensing.

[11]  Edward H. Adelson,et al.  Shiftable multiscale transforms , 1992, IEEE Trans. Inf. Theory.

[12]  B. N. Chatterji,et al.  An FFT-based technique for translation, rotation, and scale-invariant image registration , 1996, IEEE Trans. Image Process..

[13]  Naoki Saito,et al.  Multiresolution representations using the autocorrelation functions of compactly supported wavelets , 1993, IEEE Trans. Signal Process..

[14]  J. L. Moigne.  Parallel registration of multisensor remotely sensed imagery using wavelet coefficients , 1994 .

[15]  B. J. Devereux,et al.  Geometric correction of airborne scanner imagery by matching Delaunay triangles , 1990 .

[16]  Ernest M. Stokely,et al.  Laplacian and orthogonal wavelet pyramid decompositions in coarse-to-fine registration , 1993, IEEE Trans. Signal Process..

[17]  Israel Cohen,et al.  Shift invariant wavelet packet bases , 1995, 1995 International Conference on Acoustics, Speech, and Signal Processing.

[18]  Marco Corvi,et al.  Multiresolution image registration , 1995, Proceedings., International Conference on Image Processing.

[19]  Ronald W. Schafer,et al.  Template matching based on a grayscale hit-or-miss transform , 1996, IEEE Trans. Image Process..

[20]  R. Wong,et al.  Scene matching with invariant moments , 1978 .

[21]  T. W. Parks,et al.  Translation invariant wavelet transforms with symmetric extensions , 1996, 1996 IEEE Digital Signal Processing Workshop Proceedings.

[22]  A. Bijaoui,et al.  Geometrical registration of images: the multiresolution approach , 1993 .

[23]  David Casasent,et al.  Optical correlation filter fusion for object detection , 1994 .

[24]  J. Le Moigne,et al.  Towards a parallel registration of multiple resolution remote sensing data , 1995 .

[25]  C. D. Kuglin,et al.  The phase correlation image alignment method , 1975 .