Early detection of intracranial hemorrhage is crucial to saving patients' neurological function and even life, and accurate quantification of hematoma volume will provide an essential basis for clinical decision-making. Non-contrast computed tomography is a standard imaging method for intracranial hemorrhage. With the continuous development of artificial intelligence, an increasing number of studies are applying artificial intelligence to the early detection and segmentation of non-contrast CT images of intracranial hemorrhage. This article reviews the recent research progress of artificial intelligence in the detection, subtype classification, and hematoma segmentation of intracranial hemorrhage, in order to verify whether artificial intelligence can construct an accurate automatic classification system for intracranial hemorrhage to reduce the error rate, and to provide a basis for assisting clinicians in making accurate diagnosis and treatment plans.