Iris-Biometric Fuzzy Commitment Schemes under Image Compression

摘要: With the introduction of template protection techniques, privacy and security biometric data have been enforced. Meeting required properties irreversibility, i.e. avoiding a reconstruction original features, unlinkability among each other, can enhance existing systems in case tokens are stolen. However, with increasing resolution number enrolled users systems, means to compress signals become an imminent need practice, raising questions about impact image compression on recognition accuracy schemes, which particularly sensitive any sort signal degradation. This paper addresses important topic iris-biometric fuzzy commitment schemes’ robustness respect noise. Experiments using scheme indicate, that medium does not drastically effect key retrieval performance.