Breast cancer is one of the most important causes of increased women death rate in the world. Mammography is the most efficient approach for the early identification of breast diseases. The major objective of mammography is to identify small, non-palpable cancers during its premature stage. On the other hand, mammograms are extremely complicated to interpret being the fact that the pathological transformations of the breast are slight and their visibility is very poor with low contrast and noise. Mammograms have the valuable information such as microcalcifications and masses, which are extremely complicated to identify because mammograms are of low-contrast. Since the mammogram images are very noisy, low-contrast, blur and fuzzy, it is necessary to enhance the mammogram images for accurate identification and early diagnosis of breast cancer. In this paper, proposed an efficient technique to enhance the mammogram image using various transforms. The various transforms are wavelet transform, Curvelet transform, contourlet transform, Nonsubsampled transform. The drawback of wavelet transform is the method in which problem of filling missing data will occur. In Curvelet method the disadvantage is poor directional specificity of the images. In contourlet transform the image enhancement cannot capture the geometric information of images and tend to amplify noises when they are applied to noisy images since they cannot distinguish noises from weak edges. This entire drawback is overcome by the Nonsubsampled Contourlet transform. In order to determine the effectiveness of the proposed technique, experiments were carried using two UCI machine learning dataset and evaluated based on the PSNR value and MSE.

Mammogram Images Wavelet Transform Curvelet Transform Contourlet Transform Non Subsampled Transform and Image Enhancement