A log-ratio based unsharp masking (UM) approach for enhancement of digital mammograms
Siddharth, Rohit Gupta, Vikrant Bhateja · 2012
In digital mammographic images the characteristics of lesions get enhanced upon the application of conventional Unsharp Masking (UM) algorithms, but on doing so the pixels go out of range; this constraint can be handled if rescaling process is performed on the enhanced images. The proposed UM approach combines nonlinear enhancement operator (using sigmoid) and region segmentation using log-ratio operators. Unlike linear operators, log-ratio operators eliminate the need of scaling process to be applied on the enhanced images. Experimental results show that the proposed approach offers potential advantage in medical applications by enhancing lesion details in digital mammographic images.