Computer-aided detection and diagnosis systems have been the focus by a great number of endeavor researchers, particularly in detecting, diagnosing, and classifying brain hemorrhages. In this paper, we propose a system, which can automatically identify and classify the existence of brain hemorrhages. Our proposed method emphasizes on analyzing brain hemorrhage regions from medical images. It includes six stages: determining Hounsfield units, processing image segmentation, extracting the brain hemorrhage regions, extracting features and classifying brain hemorrhages, estimating the timing of hemorrhage. Our experimental results show that the accuracy of detection of brain hemorrhages is 100% and the classification of brain hemorrhages achieves the accuracy of 95.3%. In addition, our method also determines the bleeding timing to assist doctors with timely treatment.