Digital Forensics for Skulls Classification in Physical Anthropology Collection Management

Digital forensics plays a crucial role in the classification of skulls within physical anthropology collection management, streamlining the identification and organization of skeletal remains. This study presents a system for the classification of human skulls based on the presence or absence of a mandible, using advanced image processing techniques. A dataset comprising images of human skulls is utilized, where each image is subjected to a series of preprocessing steps to enhance its quality and facilitate feature extraction. The system employs various feature extraction methods, including Gray-Level Co-occurrence Matrix (GLCM), Discrete Wavelet Transform (DWT), Gabor features, and Segmentation-based Fractal Texture Analysis (SFTA), to capture the intricate details of the skull images. These features are critical for distinguishing skulls with mandibles from those without. A Support Vector Machine (SVM) classifier is trained on the extracted features, using labeled data to learn the characteristics of each class. The classification results indicate whether a skull has a mandible or not, with high accuracy. This system provides an efficient and accurate method for skull classification, supporting the efforts of anthropologists and forensic scientists in managing large collections of skeletal remains. The implementation of this approach enhances the accuracy of skull identification, contributing to the overall efficiency and reliability of physical anthropology collection management. The study achieved promising results, with high classification accuracy, highlighting the potential of digital forensics and image processing in anthropological research.

Keywords: Skull Dataset, Machine Learning, Support Vector Machine, Image Processing Techniques and accuracy.