A Novel and Efficient Lifting Scheme based Super Resolution Reconstruction for Early Detection of Cancer in Low Resolution Mammogram Images

Mammography is the most effective method for early detection of breast diseases. However, thetypical diagnostic signs, such as masses and microcalcifications, are difficult to be detectedbecause mammograms are low contrast and noisy images. We concentrate on a special case ofsuper resolution reconstruction for early detection of cancer from low resolution mammogramimages. Super resolution reconstruction is the process of combining several low resolutionimages into a single higher resolution image. This paper describes a novel approach forenhancing the resolution of mammographic images. We are proposing an efficient lifting waveletbased denoising with adaptive interpolation for super resolution reconstruction. Under this framework, the digitized low resolution mammographic images are decomposed into many levels toobtain different frequency bands. We use Daubechies (D4) lifting schemes to decompose lowresolution mammogram images into multilevel scale and wavelet coefficients. Then our proposednovel soft thresholding technique is used to remove the noisy coefficients, by fixing optimumthreshold value. In order to obtain an image of higher resolution adaptive interpolation is applied.Our proposed lifting wavelet transform based restoration and adaptive interpolation preserves theedges as well as smoothens the image without introducing artifacts. The proposed algorithmavoids the application of iterative method, reduces the complexity of calculation and applies tolarge dimension low-resolution images. Experimental results show that the proposed approachhas succeeded in obtaining a high-resolution mammogram image with a high PSNR, ISNR ratioand a good visual quality.