ADVIT

ADVIT performs automated analysis of hand dorsum vein patterns to identify individuals for forensic investigations.

Key Features:

• Automated Identification Process: Automates human identification by analyzing hand dorsum vein images for feature calculation.
• Image Preprocessing and Noise Reduction: Performs preprocessing including noise removal to eliminate extraneous details such as hair and improve vein visibility.
• Vein Skeleton Extraction: Extracts vein skeletons from preprocessed images producing binary representations that highlight vein structure.
• Feature Computation and Analysis: Computes two distinct types of features from vein patterns for use in experimental evaluation.
• Advanced Machine Learning Integration: Merges hand-crafted grid-based features with deep representations derived from RESNET-50.

Scientific Applications:

• Forensic human identification: Identifies individuals from hand dorsum vein patterns as a non-invasive biometric alternative to fingerprints, DNA profiling, retina scans, and facial recognition.
• Biometric research and comparison: Enables evaluation and comparison of vein-based identification methods against other biometric modalities.

Methodology:

Hand dorsum images were collected from 50 participants; images underwent preprocessing including noise removal; vein skeletons were extracted; two distinct feature types were computed; three experimental setups evaluated identification performance; hand-crafted grid-based features were merged with deep representations from RESNET-50; reported evaluation metrics were sensitivity 0.8803, specificity 0.8890, precision 0.8849, false positive rate 0.1074, accuracy 0.8861, F1 score 0.8817, and Matthews Correlation Coefficient (MCC) 0.7636.