A Novel IHS-GA Fusion Method Based on Enhancement Vegetated Area

Pan sharpening methods aim to produce a more informative image containing the positive aspects of both source images. However, the pan sharpening process usually introduces some spectral and spatial distortions in the resulting fused image. The amount of these distortions varies highly depending on the pan sharpening technique as well as the type of data. Among the existing pan sharpening methods, the Intensity-Hue-Saturation (IHS) technique is the most widely used for its efficiency and high spatial resolution. This method converts a color image from RGB space to the IHS color space. In next step the I (intensity) band is replaced by the panchromatic image. Before fusing the images, a histogram matching is performed on the multispectral and the panchromatic image. When the IHS method is used for IKONOS or QuickBird imagery, there is a significant color distortion which is mainly due to the wavelengths range of the panchromatic image. Regarding the fact that in the green vegetated regions panchromatic gray values are much larger than the gray values of intensity image. A novel method is proposed which spatially adjusts the intensity image in vegetated areas. To do so the normalized difference vegetation index (NDVI) is used to identify vegetation areas where the green band is enhanced according to the relation of red and NIR bands. In this paper vegetated areas are first found via thresholding on the NDVI and then green band is enhanced in these areas. In this way an intensity image is obtained in which the gray values are comparable to the panchromatic image. Intensity image is produced as a linear combination of MS bands and, thus, its weight parameters have a direct effect on the final fusion result. In the proposed method weight parameters are estimated by Genetic Algorithm. Beside the genetic optimization algorithm is used to find the best optimum weight parameters in order to gain the best intensity image. Visual interpretation and statistical analysis proved the efficiency of the proposed method as it significantly improved the fusion quality in comparison to conventional IHS technique. Spatial quality can be judged visually, but color changes more difficult to be recognized in this manner The spectral quality of pan-sharpened images was determined according to the changes in colors of the fused images as compared to the MS reference images. The accuracy of the proposed pan sharpening technique was also evaluated in terms of different spatial and spectral metrics. In this study, 7 metrics (Correlation Coefficient, ERGAS, RASE, RMSE, SAM, SID and Spatial Coefficient) have been used in order to determine the quality of the pan-sharpened images. Experiments were conducted on three different high resolution data sets obtained by three different imaging sensors, IKONOS, QuickBird and WorldWiew-2. The IHS pan-sharpening method results good spatial quality and is a commonly used algorithm for its speed and simplicity. The result of this proposed method showed that the evaluation metrics are more promising for our fused image in comparison to other pan sharpening methods. In fact, A combination of enhanced vegetated areas, genetic algorithm optimization and integration of IHS improves the spectral and spatial quality.