Some Advanced Techniques for spot 4 XI Data Handling
IV. A New Color Composite Model for Spot 4 HRVIR Data
SPOT 4 XI with 4 spectral bands provides 24 different color composites using the RGB model. Each RGB color composite enhances certain land cover characteristics. However, none of them is capable to display information available in all 4 spectral bands. The authors have conducted an experiment to develop a color composite using all 4 spectral bands. This new color composite is based on data transformation from 4 dimensional conic vector space into 3 dimensional orthogonal space.
In general, there is possibility to transform data from n to 3 dimension space. Some degradation of data quality, of course, can be found in the result, however, experiments have confirmed that the visualized transformed data show more information than any of the conventional three band color composites. The transformation can be made using the following equation:
Where pi is original image digital count and p'i is transformed value. The coefficients a1, an, b1, bn, c1, cn can be computed using different transformation model. In this case the authors used 4 dimensional conic vector space to transform data from 4 to 3 dimension space. For the case of SPOT 4 data the transformation is done by the following equation:
Because the transformed components are in achromatic space so it is necessary to convert them to IHS and RGB space for color visualization. The conversion can be done by any of common HIS-RGB algorithms. The new color composite provides more information than any of the conventional ones. The visualized image is an excellent tool for vegetation study and water and infrastructure mapping. Conversion of transformed components p'i into I,H,S system is done by formulas:
On the Figure 3 is color composite created by this approach. This conversion has been applied for all pixel vectors in the image. Absolute calibration could be applied to ensure stability of the output color. To obtain specifically desired color, some offset of H could be added. When comparing this image with color composite on Figure 1 we could see that the new color composite is much more better than the standard one. When a composite is made by assigning component 1 to blue, component 2 to green and component 3 to red color respectively, vegetation is displayed always in green,
water in blue like in true color mode. Therefore the authors has named it as quasi-true color composite. Because of existence of the SWIR band which is not much impacted by atmospheric water vapor and aerosol so the final image is much more clear with higher contrast than the conventional one. Many land cover types such as urban, turbid water, bare soil that have similar color in standard color composite are very easy to be recognized each from other in the new color composite.
V. Conclusion
From this research we could make some conclusions:
- The SPOT 4 XI data with new SWIR band is excellent information source for land cover mapping and environmental research.
- Some saturation is found out in the SWIR band for cloud and bright ground objects. This occurs mostly for image data received in high gain mode.
- The graphical analysis of spectral reflectance curve (GASC) algorithm can be applied for automated classification of SPOT 4 XI data.
- Due to different gain mode of SPOT 4 data, absolute calibration should be applied before classification and image invariant used for digital description of land cover must be computed using absolutely calibrated pixel vector.
- It is possible to create new color composite using all four SPOT 4 XI bands by transformation matrix given in the paper. The visualized image provides more information than the conventional standard color composite and enhances many land cover objects. The new color composite is suitable for vegetation study, water body and infrastructure mapping.
Acknowledgement
The authors would like to acknowledge the Satellite Remote Sensing Laboratory, NCU of Taiwan for providing SPOT 4 data. The authors also thank the Fundamental Research Programme of Vietnam for funding the research.
Reference
-
SPOT IMAGE: The SPOT Scene Standard Digital Product Format S4-ST-73=01-SI
- Nguyen Dinh Duong. Graphical Analysis of Spectral Reflectance Curve. Proceedings of the 18th Asian Conference on Remote Sensing. 20 - 24 October 1997, Kualalumpur Malaysia.
- Nguyen Dinh Duong. Total Reflected Radiance Index- An Index to Support Land Cover Classification. Proceedings of the 19th Asian Conference on Remote Sensing. 16 - 20 November 1998, Manila, Philippines
- Nguyen Dinh Duong. Land Cover Category Definition by Image Invariants for Automated Classification. International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7/3, Commission VII. ISPRS 2000 Amsterdam, the Netherlands.
