| Summary: | Digital mammography proved its efficacy in the diagnosis of breast cancer as an adequate and easy tool in detection tumors in their early stages. Mammograms have useful information on cancer symptoms such as micro calcifications and masses, which are difficult to identify because mammograms images suffer from some defects such as high noise, low-contrast, blur and fuzzy. In addition, mammography has major problem due to high breast density that obscures the mammographic image leading to more difficulty in differentiating between normal dense tissue and cancerous tissue. Therefore, for accurate identification and early diagnosis of breast cancer, mammograms images must be enhanced. Image enhancement commonly focuses on enhancing image details and removing noises. Using image processing techniques for mammogram images help to differentiate a special data that contain specific features of the tumors, which could be helpful in classifying benign and malignant tumors. This research focuses on salt and pepper noise remove and image enhancement to increase the mammography quality and improve early breast cancer detection. To achieve this purpose, a special technique is used that includes two stages image denoising base filtering and one stage for contrast enhancement. The filtering stages include the using of median and wiener filters. The contrast enhancement stage uses contrast limited adaptive histogram equalization (CLAHE). The evaluation of the performance is measured by PSNF and MSE for the filters and by contrast histogram for the CLAHE. The results show better performance of the research technique compared with other methods in terms of high PSNR(47.4750) and low MSE(1.1630). For future work, the technique will be evaluated with other type of noise.
|
|---|
