The Next Step in Biometric Data Fusion Ged Griffin APCOA Melbourne 23 February 2011
Biometrics
Types of Systems Singular Multiple singular Multi-modal
Context Security Homeland security/border protection Policing/investigative activity
Reasons for Use Identity Verification Identification of person Person in custody Watchlist Human remains Crime scene
Typical Operation Input/ Capture Signal Processing Output/ Decision Template Creation/ Storage Template Search/ Comparison
Sensor Iris Retina DNA Fac. Rec. F/Print Extraction Iris Retina DNA Fac. Rec. F/Print Matching Iris Retina DNA Fac. Rec. F/Print Decision Biometric Data Fusion
Biometric Maturity
Considerations Maturity of biometric products (esp. multimodal) Next Generation Identification (NGI) Impact of transnational organised crime Convergence of biometrics and forensic science Application of multi-media data fusion techniques on biometrics and forensic science Improved capabilities through advancements in information and communications technology (Cloud computing & quantum computing) Proliferation of mobile devices “Enter once, search many” operation
Policing Investigative Activity What is an investigation? “an investigation is a search for the truth, in the interests of justice and in accordance with the specifications of the law” What does an investigator need to establish 1.Has a crime been committed 2.Identity of the offender(s) 3.Location of the offender(s) 4.Connection between the offender and the offence
Crime Scene Primary Secondary Locard’s principle of exchange, ‘every contact leaves its trace’ Evidence located Searchable Identifiable Probative weight Intelligence value
Biometrics Personal Physical Behavioural Environmental Bioinformatics Commodity Bioinformatics Data Fusion Decision/ Action Linking Biometrics and Forensic Science
Bioinformatics National DNA databases for plants, animals, birds, and insects Geo-referenced templates may help to locate unique geographical areas Help in the management of animal/plant disease and bio- containment strategies Identify stolen animals and stock Identify owners in animal welfare investigation Identify owners of animals in criminal investigations and local enforcement (animal hair at crime scenes, dog attacks and fecial deposits in public spaces)
Commodity - Drugs Organic drugs - DNA analysis Synthetic drugs chemical composition Pattern recognition of images on packaging Automated extraction and matching of pill press impressions/features
Commodity – Tool Marks Tool mark examination and firearm/ballistic examination Develop templates for common articles (eg. shoe impressions) Reactive or Proactive Capture and storage of images of known and recovered firearms
The future Forensic multi-media data fusion and information retrieval platform Consideration of evidence in the context of all other evidence Open architecture to support interoperability Integrated/networked computerised system(s) Facilitate high volumes of transactions and fast turn around Processing scalability and matcher performance scalability
Ged Griffin Mb: