EVALUATION OF SATELLITE IMAGE FUSION USING WAVELET TRANSFORM 
Oguz Gungor Jie Shan 
Geomatics Engineering, School of Civil Engineering, Purdue University 
550 Stadium Mall Drive, West Lafayette, IN 47907-2051, USA — (ogungor, jshan) @purdue.edu 
Commission III, WG IIV6 
Key Words: Fusion, Remote Sensing, Transformation, Multispectral, Spatial, Spectral, Algorithms, Resolution. 
ABSTRACT: 
This paper addresses the principles, implementation and evaluation of wavelet transformation based image fusion. 2-D discrete 
wavelet transformation is presented concisely to facility the understanding of the wavelet based image fusion method. To best retain 
the quality of the input images, we propose a strategy that minimizes the necessary resampling operations to limit potential image 
quality deterioration. In the proposed fusion approach, the wavelet coefficients for the fused images are selected based on the 
suggested maximum magnitude criterion. To evaluate the outcome images, other popular fusion methods including principal 
component transformation, Brovey and multiplicative transformation approaches are applied to the same images and the results are 
compared to the ones from the wavelet based approach. Fusion results are evaluated both visually and numerically. A quality matrix 
is calculated based on the correlation coefficients between the fused image and the original image. It is shown that this quality 
measure can indicate the information content of the fused image comparing to the input panchromatic and multispectral images. Our 
results clearly suggest that the wavelet based fusion can yield superior properties to other existing methods in terms of both spatial 
and spectral resolutions, and their visual appearance. This study is carried out using multiple images over the Davis-Purdue 
Agricultural Center (DPAC) and its vicinity with both urban and rural features. Images used include QuickBird panchromatic band 
(0.7 m) and multispectral bands (2.7m), and Ikonos panchromatic (1 m) and multispectral bands (4 m). 
Principle Component Analysis (PCA), Brovey and 
Multiplicative Transformation methods to evaluate the 
proposed wavelet transformation approach. Finally, a 
1. INTRODUCTION 
Image fusion, in. general, can be described as a process of quantitative evaluation criterion is proposed to evaluate the 
producing a single image from two or more images that are quality of the fusion outcome. 
collected from the same or different sensors. The objective of 
the fusion process is to keep maximum spectral information 
from the original multispectral image while increasing the 
spatial resolution (Chavez et al, 1991; Ranchin et al, 2003). 
2. WAVELETS AND WAVELET TRANSFORM 
Military, medical imaging, computer vision, robotic industry In this paper, 2-D Discrete Wavelet Transform (DWT) is used 
and remote sensing are some of the fields benefiting from the for image fusion process. Wavelet transform is defined as the 
image fusion. sum over all time of the signal multiplied by scaled, shifted 
version of the mother wavelet y(t). Similar to the Fourier 
In the field of remote sensing, lower spatial resolution 
is ral images need to be fused with higher resolution A ; : 
multispectral images need to b 8 different frequencies, wavelet transform decomposes a signal 
panchromatie masses: The fusion techniques should ensure that into the scaled and/or shifted versions of the mother wavelet 
all important spatial and spectral information in the input (Misiti, 2002) 
images is transferred into the fused image, without introducing 
artifacts or inconsistencies, which may damage the quality of 
the fused image and distract or mislead the human observer. 
Furthermore, in the fused image irrelevant features and noise 
should be suppressed to a maximum extent. Image fusion can 
be performed at pixel, feature and decision levels according to 272 i 
the stage at which the fusion takes place (Pohl and van V un (1) =2 vl t= n) (1) 
Genderen, 1998). 
analysis that breaks a signal into different sine waves of 
In DWT, instead of calculating wavelet coefficients at every 
possible scale, the scales and shifts are usually based on power 
of two. If we have a mother wavelet, 
In this study, a pixel level multispectral image fusion process A signal (1) can be expressed by wavelets as 
using wavelet transform approach is performed. The fusion 
process is implemented to two categories: images collected by f (= > Con mon (1) (2) 
the same sensors at the same time, and images collected by m,n 
different sensors. For the same sensors, a QuickBird where m and n are integers (Nikolov et al, 2001). Here, 
panchromatic image is fused with QuickBird multispectral PAGS is the dilated and/or translated version of the mother 
images. For the different sensors, a QuickBird panchromatic 
image is fused with Ikonos multispectral images. Haar and 
Daubechies (DB) wavelets are used in this study. For coefficients are needed., These coefficients denote the 
comparison purpose, the same images are also fused using approximation of f at each scale. For example a,,, and a,,,,, 
wavelet y/ . To implement an iterative wavelet transform dA, 
1244