COLOR CORRECTION OF FUSED IMAGES USING COLOR SPACE 
TRANSFORMATION 
Tsukasa Hosomura and Kouichi Katou 
Kanazawa Institute of Technology, Japan 
Commission VII, Working Group 1 
KEY WORDS: Remote Sensing, Technology, Fusion, Algorithm, Multispectral, Method 
ABSTRACT: 
Several techniques have been developed to merge SPOT 10-m resolution panchromatic data with SPOT 20-m 
resolution multispectral data or Landsat TM 30-m resolution multispectral data. That is the effectively 10-m resolution 
multispectral images produced through the various merging methods contain the high resolution information of the 
respective panchromatic images while maintaining the basic color content of the original multispectral data. The utility 
of hue-saturation-intensity (HSI) transformation procedures for creating such composite image. For solving this 
problem, a method of color correction by transformed color space was developed. The method is simple and easy to 
implementation. Only intensity was changed. Hue and saturation was kept previous value. Two kind of points are set in 
the color space. One is Fixed point which is fixed in the transformation. Another is Reference point which is moved to 
target point in the transformation. Using this method, we can correct plural colors independently in a image. Dark 
forest area which were obtained by using HSI transform algorithm became bright. Forest area in obtained results was 
equivalent to the value of the same area in the original color image. This method is very useful for making multi-sensor 
fused images clear 
groups. These method basically use RGB--HSI 
transformation. Where RGB means red, green and blue 
INTRODUCTION image and HSI means hue, saturation and intensity 
image. 
Many techniques have been proposed to merge SPOT (1) RGB--HSI transformation. I is replaced by 
10-m resolution panchromatic data with SPOT 20-m panchromatic image. HSI--RGB inverse 
resolution multispectral data or Landsat TM 30-m transformation. 
resolution multispectral data. The fused images arc (2) RGB--HSI transformation. I is added the high 
very useful for land cover classification . land use frequency components. HSI--RGB inverse 
mapping and other wide application fields. However, transformation. 
there are several problems when multi sensor image 
fusion are tried. In order to solve these problems color By using first method, obtained fused image contains 
adjust method was introduced in this research. high resolution information which come from 
panchromatic image. But density level of obtained 
fused image is different from that of original color 
DESCRIPTION OF METHOD image. Especially, this problem was caused when color 
composite image contains the near infrared band. 
The flow chart of this research is shown in Fig.l. Because panchromatic image don't has the information 
Existing methods can be divided into following 2 of near infrared band. 
304 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996