Combination of Wavelet Analysis and Non-Linear Enhancement Function for Computer Aided Detection of Microcalcifications

Microcalcifications are small densities that appear as bright spots on mammograms. Its detection plays a very important role in breast cancer detection at early stages in digitized mammographic images. This paper presents an approach for computer aided detection of microcalcifications. The main idea of the proposed approach is to apply wavelet analysis followed by a non-linear transformation function for contrast enhancement. The image is decomposed into sub-bands using Daubechies-6 orthogonal wavelet filter. With this, the lowfrequency sub-band is suppressed and then the image is reconstructed from the high frequency sub-bands. Contrast enhancement is carried out using a linear combination of double sigmoid function, which is objectively evaluated using PSNR & a