Middle finger bone assessment for age identification

People who died because of natural disaster, airplane crash or vehicle accident are hard to identified. There are many parameters that used by forensic doctor to identify victims corpes. Bone is one of parameters that used by doctors to identify victims age, it provide accurate result compared to other diagnoses. But, it takes a long time for the doctors to identify manually. From those problems, we tried to develop an automatic system which can identify victims age using middle finger bone. An active shape model segmentation method applied in this system to extract middle finger bone. There are six parts of middle finger bone that used to analize age, proximal epyphisis, proximal metaphysis, middle epyphysis, middle metaphysis, distal epyphysis, and distal metaphysis. We measured the length of each parts to be input in age classification using k-Nearest Neighbor method. By using this method, 85% from 73 different experimental data has been succeeded to identified. We believe this can bring benefit for the future of forensic identification.