METHODS

1. Image Fusion Techniques

Five best image fusion techniques based on literature review and comparative evaluation were used. These techniques are smoothing Filter- SFIM (Bharath et al., 2009), COS (Component Substitution) (Kumar et al., 2011), High Pass Fusion-HPF (Kumar et al., 2009a), High Pass Filter-HPF (Kumar et al., 2009b) and High Pass Modulation-HPM (Kumar et al., 2009b).

2. Methods of validation

The performance of the image fusion techniques were analysed qualitatively and quantitatively by visual interpretation and correlation coefficient (CC) that is often used as a similarity metric in image fusion. However, CC is insensitive to a constant gain and bias between two images and does not allow subtle discrimination of possible fusion artifacts (Aiazzi et al., 2002). In addition, a universal image quality index (UIQI) (Wang et al., 2005) is used to measure the similarity between two images. UIQI is designed by modelling any image distortion as a combination of three factors: loss of correlation, radiometric distortion, and contrast distortion given by:

                                                                                         (1)

The first component is the CC for A (original MS band) and B (fused MS band). The second component measures how close the mean gray levels of A and B is, while the third measures similarity between the contrasts of A and B. The dynamic range is [-1, 1]. If two images are identical, the similarity is maximal and equals 1. In addition, minimum (min), maximum (max), and sd of the original and fused bands were also analysed. These are the most commonly used indices/measures found in literature that provide robust statistics for validating the fused images with the original reference images at the original image resolution.

3. Pattern classifier

A pattern classifier (K-Means Clustering) was used to delineate the wetlands (Ramachandra and Kumar, 2008) from the fused spectral bands for the different sensor data.