This chapter discusses lessons that can be learned in biometrics from the field of the forensic sciences. It acknowledges the fact that biometrics and forensics are both very old research disciplines which have a very different perspective in life: While work in biometrics is mostly focused on application issues, like achieving certain error levels, forensics need a very thorough backing to achieve the ultimate goal in this field, admissability in court. This automatically results in high standards for methods that exceed simple performance issues by far. One aspect that is used in this chapter as the focus of the discussions is the matter of privacy. In the first half of the chapter it is illustrated by example how current research work in one digitized forensics field, here digitised dactyloscopy (i.e. the science of forensic analysis of fingerprint traces), influences the current view on fingerprint biometrics and which lessons in regards to privacy can be derived. In the second half, the ever popular field of face biometrics is addressed as an example of an widely used biometric modality in desperate need of not only digital image forensics but also guidelines for privacy preserving methods.

Chapter Contents:

• 3.1 Introduction: forensic science and selected privacy concepts
• 3.2 Privacy concepts — findings from digitised forensics of latent fingerprints
• 3.2.1 Sensor-acquisition-related privacy-preserving guidelines
• 3.2.1.1 Localisation of fingerprints using a coarse scan: findings for privacy-preserving dot distances and resolutions
• 3.2.1.2 Small area scans on the example of age detection: findings for privacy-preserving spatial dimensions
• 3.2.1.3 Structural and semantical selection: findings for privacy-preserving sample part selection
• 3.2.2 Privacy-preserving-benchmarking concepts and guidelines
• 3.2.2.1 Privacy-preserving data sets for benchmarking (testing): artificial sweat-printed computer-generated fingerprints
• 3.2.2.2 Simulation framework: from StirMark to StirTrace for artefact simulations in forensics and biometrics
• 3.3 Privacy concepts — findings from digital forensics of face-morphing detection in face authentication systems
• 3.3.1 Face-morphing attacks — generalised attack procedure and privacy implications
• 3.3.2 Media forensic investigations and biometrics on the example of face-morphing attacks
• Acknowledgements
• References

Inspec keywords: face recognition; image forensics; data privacy; fingerprint identification

Other keywords: fingerprint biometrics; privacy preserving methods; digitised dactyloscopy; digitized forensics field; biometric modality; forensic analysis; digital image forensics; face biometrics; fingerprint traces; privacy concepts; forensic sciences

Subjects: Optical, image and video signal processing; Computer vision and image processing techniques; Data security