In: Wagner W., Szekely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
i 
681 
algorithms combing wavelet transform and IHS transform or 
PC A transform. Recently, some authors introduce sensors’ 
spectral response and ground spectral features into fusion 
technology on the basis of previous three categories, and 
other authors utilize the regularization method to optimize 
the last fusion results so as to satisfy the higher resolution 
multispectral image model. 
2. COMPONENT SUBSTITUTION 
FUSION TECHNIQUE 
In general, the Component Substitution fusion Technique 
consists of three steps: Step 1, Forward transform is applied 
to the multispectral bands after they have been registered to 
the panchromatic band; Step 2, A component of the new data 
space similar to the panchromatic band is replaced with the 
higher resolution band; Step 3, The fused results are finally 
obtained via inverse transform to the original space. 
The typical algorithms of component substitution fusion 
technique are IHS transform fusion algorithm (Carper, 1990, 
Shettigara, 1992, Chavez, 1991). This algorithm is suitable 
when exactly three multispectral (MS) bands are concerned 
since the IHS transform is defined for three components only. 
Usually, Panchromatic band (PAN) is histogram-matched, 
i.e., radiometrically transformed by a constant gain and bias 
in such a way that it exhibits mean and variance that are the 
same as Intensity, before substitution is carried out. When 
more than three bands are available, Tu et al (Tu, et al, 2004) 
present a generalized IHS (GIHS) transform by including the 
response of the near-infrared (NIR) band into the intensity 
component. The GIHS-GA (Garzelli and Nencini, 2006) is 
based on CS strategy and genetic algorithm. The weights of 
the MS bands in synthesizing the intensity component and 
the injection gains are achieved by minimizing a global 
distortion metrics (Q4, in this case) by means of a GA. The 
GIHS-TP (Choi, 2006) is a CS-based method that trades off 
the performances of GIHS in terms of spectral distortion and 
spatial enhancement. Aiazzi et al (Aiazzi, et al, 2007) 
introduce multivariate regression to improve spectral quality. 
In the method based on IHS, a generalized intensity 
component is defined as the weighted average of the MS 
bands. The weights are obtained as regression coefficients 
between the MS bands and the spatially degraded PAN 
image with the aim of capturing the spectral responses of the 
sensors. Gonzales Audicana and Otazu (Gonzales Audicana 
and Otazu, et al, 2006) present a low computational-cost 
method to fuse IKONOS images using the spectral response 
function of its sensors. Andreja and Kris v tof (Andreja and O. 
Kris“tof, 2006) found that for preserving spectral 
characteristics, high level of similarity between the 
panchromatic image and the respective multispectral 
intensity is needed. In order to preserve spectral and spatial 
resolution, spectral sensitivity of multispectral and 
panchromatic data was performed, and digital values in 
individual bands have been modified before fusion. Malpica 
(Malpica, 2007) present a technique which consists of a hue 
spectral adjustment scheme integrated with an 
intensity-hue saturation transformation for vegetation 
enhancement. Ling and Ehlers, et al (Ling, Ehlers, et al, 
2007) present a method which combines a standard IHS 
transform with FFT filtering of both the panchromatic image 
and the intensity component of the original multispectral 
image. 
Other common used CS-based method, PCA transform, 
(Shettigara, 1992, Chavez, 1991) make an assumption that 
the first principal component (PC) of high variance is an 
ideal choice for replacing or injecting it with high spatial 
details from the highresolution histogram-matched PAN 
image. Shah, et al (Shah, Younan, and King, 2008) use the 
adaptive PCA to reduce the spectral distortion in the fusion 
scheme combining adaptive PCA approach and contourlets. 
Another CS technique reported in the literature is 
Gram-Schmidt (GS) spectral sharpening (Laben and Brower, 
2000), which is widely used since it has been implemented 
in the Environment for Visualizing Images (ENVI) program 
package. Aiazzi et al (Aiazzi, et al, 2007) adopts multivariate 
regression to create the synthetic low-resolution-intensity 
images which is used in the Gram-Schmidt transform. The 
proposed enhanced strategy is effective in improving the 
quality of the images than ordinary GS technique. 
UNB-pansharp (Zhang, 2002) algorithm developed at the 
UNB, Canada, is based on CS. The least squares technique is 
utilized to reduce color distortion, by identifying the best fit 
between gray values of individual image bands and adjusting 
the contribution of the individual bands to the fusion result. 
3. MODULATION-BASED FUSION