.png){kind=icon}
.png){kind=icon}
Return
A Novel Image Registration Algorithm Using Wavelet Transform and Matrix-Multiply Discrete Fourier Transform
.png){kind=icon}
PRE
AI
DOI：10.1109/LGRS.2020.3031335.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Nearly all existing image registration algorithms are based on the full information (e.g., intensity and/or features) carried by the images being registered. Such a full-image-based strategy can achieve high-precision registration (e.g., at the subpixel level) but at the expense of long computation times, especially for high spatial resolution remote-sensing (RS) images. This letter presents a novel subimage-based registration method that is expected to dramatically reduce the time consumption but still meet subpixel accuracy requirements. The core idea behind this method is that the transform pattern of the reference and sensed images under a global transformation is theoretically similar to that of subimages derived from the full image, and vice versa. Hence, this method first divides the original reference image into subimages with nearly equal sizes, and extracts the subimage with the maximum wavelet coefficient summation to serve as a new reference image using the wavelet transform technique. Then, the two-step methods based on the phase correlation and matrix-multiply discrete Fourier transform (DFT) algorithms are applied to register the sensed image to the reference subimage. Finally, the proposed method is tested with simulated data sets. The results indicate that this method can perform subpixel registration as achieved by full-image-based methods but with the benefit of much shorter computation times (e.g., approximately 89x0025; compared with a widely used rapid registration algorithm named enhanced phase correlation).
Keywords:
Image registration
Discrete Fourier transforms
Correlation
Wavelet transforms
Data mining
Remote sensing
Discrete Fourier transform (DFT)
image registration
remote sensing (RS)
subimage
wavelet transform
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
16.4
Papers:
1.0W
Citations:
5.1K