Performance analysis of hyperspherical colour sharpening method for IRS satellite images

Abstract

Various image fusion methods have been developed and investigated for different remote sensing (RS) applications. Hyperspherical Colour Sharpening (HCS) method was recently proposed for World View-2 imagery. A limited study has been carried out to find the performance of HCS method for other datasets. In this paper, an experiment is engineered in which HCS method was applied on Indian remote sensing (IRS) datasets. The performance analysis of the method was carried by both qualitative and quantitative methods. In addition to that the quality of indices image for each method is compared to analyse the suitability of methods for various applications based on these indices. Brovey transformation (BT), principal component substitution (PCS), high pass filtering (HPF) and discrete wavelet transform-based principal component substitution (DWT-PCS) were also applied on the selected data and used in comparative analysis with the HCS method. The study reveals that HCS method outperforms in terms of the spectral fidelity, but produces some shortcoming for spatial resolutions. The quality of indices images show that BT and HCS methods do not hold the spatial details after fusion indices image computation, while indices from HPF, DWT-PCS and PCS hold some of spatial information injected into fused output.