The statistical methods based on discrete wavelet transform for remote sensing images fusion.

In remote sensing images, numerous image fusion methods have been developed to produce high-resolution multispectral images from multispectral (MS) and panchromatic (PAN) image. This work suggests suitable algorithm that maintains the spectral and spatial information content and to overcome the spectral degradation of satellite images. Three statistical methods are used to implement the fusion process based on Discrete wavelet transform (DWT). Absolute Max, Mean and Standard Deviation and Simple Linear Regression (SLR). The DWT are applied on both MS bands and PAN images to produce high and low frequency components, then the statistical methods are applied on high components as a fusion method. Lastly the inverse DWT is implemented to produce the final fused image. Three metrics (Standard Deviation SD, Peak signal-to-noise ratio PSNR and Correlation Coefficient CC) are implemented to evaluate the obtain results quantitatively in addition to visual assessment. Good qualitative and quantitative results were obtained with a Absolute Max and SLR over Mean and Standard Deviation. [ABSTRACT FROM AUTHOR]