Breast cancer is a leading cause of cancer deaths among women. Early detection is the most effective way to reduce mortality. Mammography is currently the best method for early detection. An increasing number of countries have started mass screening programs that have resulted in a large increase in the number of mammograms requiring interpretation. In the interpretation process, radiologists carefully search each image for any visual sign of abnormality. However, abnormalities are often embedded in and camouflaged by varying densities of breast tissue structures. Indeed, estimates indicate that 10 to 30 percent of breast cancers are missed by radiologists during routine screening. Limitation in the number of radiologists and large number of mammograms have caused the interpretation of mammograms costly, time consuming and even incorrect .Furthermore, the ensitivity of man's eye decreases by increasing the number of images. So, it is needed to design automatic system for breast cancer diagnosis at the first level. In computer aided diagnosis on digital mammograms, some pre-processing like breast contour extraction and pectoral muscle segmentation are performed to limit the region of interest. Then the main processing is done in order to detect suspicious details on mammograms. Some new methods for extracting breast contour, segmentation of pectoral muscle, enhancement of mammogram in order to detect microcalcifications and detecting tumor and architectural disorders are introduced in this report. Results verify the performance of these methods.