:Authentication is the process of automatically recognizing the correct person. Presently, the biometric identification systems are based on static features like face[2], iris[3], palm print[4], voice[5] and fingerprint impression of the user, which mostly remains unchanged over time. The performance of a biometric identification system is measured based on accuracy, efficiency, security, and privacy. Biometric systems are a combination of multiple sensors, multiple algorithms, and numerous instances, making it more accurate, reliable, secure, and robust[6]. Identifying an injured person is challenging in disasters like tsunamis and earthquakes and catastrophic accidents. In such cases hand radiographs may be considered. Because bones cannot be easily damaged. The existing method uses KNN (K-Nearest Neighbor) classifier[8]. It consists of training and testing stage. In training stage, CNN is applied which includes mainly convolution layer, ReLU layer and Max pooling layer. The proposed method uses CNN classifier. It consists of trainingand testing stage.