At present, Mammography is one of the most effective methods to detect early breast cancers. However, the signs of most micro-calcifications that are the early signs of malignant tumours cannot appear clearly in an inhomogeneous background because of the complicated structures in breast. As a rule, a tool of magnified glass is required to enlarge mammograph views to check the characteristics or a suspected lesion, but there is at least 15 percent or breast cancer that is still missed. Without specific feature, when doctors have to relay on clinical histology, this will causes a large number of misdiagnosis with false negative or false positive even decrease the rate of detection, furthermore, causing pain to the patients. Therefore, it is a crucial task to need a reliable and effective approach to assist the micro-calcifications detection. This paper uses a new technology to extract micro-calcifications clusters with accurate edge effects to obtain much more hidden information which can't be detected by the naked eye on mammograms in order to help the doctors in diagnosing early breast cancer. In this paper, this research combines the findings of histopathology in benign or malignant calcifications and uses the typical application examples with the contrastive analysis to do a relevant expatiation.