Adaptive image enhancement using nonsubsampled contourlet transform domain histogram matching

In this letter, we propose a novel adaptive image enhancement algorithm based on nonsubsampled contourlet transform (NSCT) coefficient histogram matching. Firstly, the original image is decomposed in the NSCT domain. Secondly, the NSCT coefficient histograms of the original image in corresponding subbands are adaptively mapped to those of the reference image via histogram matching after threshold denoising. Finally, the enhanced image is reconstructed from the modified coefficients via inverse NSCT. Experimental results demonstrate that the proposed adaptive algorithm effectively improves subtle features while suppressing noise compared with existing algorithms.