Machine learning algorithms are suggested for detecting and classifying hemorrhage regions on head CT/MRI images with high accuracy. However, most of these algorithms are not interested in the valuable characteristic of CT/MRI images, especially Hounsfield Unit values. Besides, they also only detect and classify one of types of the intracranial hemorrhage on each image. In this paper, we propose a new approach for brain hemorrhage identification using object detection algorithms like Faster R-CNN and R-FCN. The proposed approach can detect many regions of the brain hemorrhage on a CT image. The results show that the R-FCN algorithm gives better results than the Faster R-CNN algorithm on time and accuracy of identification.