Forensic Human Identification Using Fingernail Characteristic Based on Deep Learning Method

Accurate and reliable human identification is great importance in forensic investigations, particularly in situations where conventional biometric traits are unavailable, damaged, or difficult to obtain. Fingernails represent a distinctive physiological characteristic of the human hand and can provide discriminative information for personal identification. This paper proposes a convolutional neural network-based method for automated human identification using fingernail characteristic for forensic applications. In the proposed framework, fingernail regions are first extracted from hand images and subjected to appropriate preprocessing to improve image quality and reduce irrelevant information. Subsequently, a convolutional neural network model is employed to automatically learn discriminative features from the fingernail images and classify individuals according to their learned biometric patterns. The performance of the proposed approach is evaluated using a dataset of fingernail images collected by the author of the paper under controlled imaging conditions. Experimental results demonstrate that fingernail characteristic can provide effective biometric information for human identification. The presented framework provides an automated and non-invasive approach that can potentially support forensic identification and complement conventional biometric evidence in forensic investigations.