Satellite Image Resolution Enhancement Using Double Density Dual-Tree Complex Wavelet Transform and Nonlocal Means

Abstract: A Multi scale decomposition based on double density dual-tree complex wavelet transform and edge preservation is presented for SRE of the satellite images. Resolution enhancement schemes (which are not based on wavelets) suffer from the drawback of losing high frequency contents (which results in blurring). The Discrete Wavelet Transform- based (DWT) RE scheme generates artifacts (due to a DWT shift-variant property). A wavelet-domain approach based on Double Density Dual-Tree Complex Wavelet Transform (DDDTCWT) and nonlocal means is used for RE of the satellite images. A satellite input image is decomposed by DDDTCWT to obtain high-frequency subbands. The high-frequency subband and the low-Resolution (LR) input image are interpolated using the bicubic interpolator. The high frequency subbands are passed through an NLM filter to cater for the artifacts generated by DDDTCWT. The filtered high-frequency subbands and the LR input image are combined using inverse DDDTCWT to obtain a resolution-enhanced image. The simulated results will show that technique used in this process provides better accuracy rather than prior methods.

Keywords: Pre-processing, double density dual-tree complex wavelet transform (DDDTCWT) nonlocal means filter (NLM), Reconstruction, Parameter analysis.