Chapter 7: Forensic Biometrics: Iris & Facial Recognition

Forensic Biometrics: Iris & Facial Recognition categorizes biometric systems into two primary groups: physiological biometrics, which focuses on stable physical characteristics such as fingerprints, hand geometry, vein patterns, retina structures, and the highly unique human iris, and behavioral biometrics, a more experimental field that analyzes dynamic patterns including handwriting, keystroke dynamics, voice recognition, and gait analysis. The text details the universal biometric workflow, beginning with the enrollment process where raw data is captured via sensors, followed by pre-processing to normalize the data, feature extraction to identify unique points using mathematical algorithms, and finally template generation where data is converted into a searchable format for matching. Significant attention is devoted to iris recognition as a superior biometric tool due to the stability of the iris throughout a lifespan and the use of near-infrared (NIR) wavelengths to map intricate muscle patterns into a binary IrisCode, a method that is far less intrusive and more user-friendly than outdated retina scanning technology. The summary also examines the complexities of facial recognition technology, discussing appearance-based techniques like Principal Component Analysis (PCA) which utilizes Eigenfaces, and Linear Discriminant Analysis (LDA), alongside model-based approaches like Elastic Bunch Graph Matching (EBGM) that map nonlinear facial landmarks. Furthermore, the chapter outlines the significant transition from the FBI's legacy IAFIS to the advanced Next Generation Identification (NGI) system, which integrates multimodal biometrics including the Rap Back service for ongoing criminal monitoring, the Interstate Photo System (IPS) for facial searches, the National Palm Print System (NPPS) for solving cold cases using palm impressions, and the mobile Repository of Individuals of Special Concern (RISC) for rapid field identification.