Novel 2D, 3D and multimodal biometrics system
Lyon Science Transfert offers to grant patent license and looks for industrial partners.
DESCRIPTION
Biometrics is to identify individuals using their physiological characteristics (e.g. face, iris, hand vein, fingerprint, etc.) or behavioural features (e.g. voiceprints, gait, handwriting, etc.).There is few approaches which can be used as a general tool for different biometrics, enhancing the distinctiveness of each biometric representation.
The laboratory offers a novel biometrics system that can be applied for 2D, 3D and multimodal face recognition as well as hand vein recognition. Biometric representation is based on oriented gradient maps (OGM) and SIFT-based (Scale Invariant Feature Transform) matching scheme to enable biometrics-based people identification or verification. The OGMs allows to highlight the distinctiveness of human biometric images, in particular facial texture images (2D), facial range images (3D), hand vein images, fingerprints, iris, etc. The SIFT-based matching then allows to compare one person’s biometric signature with another one despite of possible scale and rotation differences.
- ADVANTAGES / NOVELTY
This system can be used as a general method for different biometric applications: 2D, 3D, multimodal face recognition as well as hand vein recognition. It can achieve better performance and provide a higher possibility to be implemented for real condition.
APPLICATIONS
Biometrics (e.g. face and hand vein) based security applications
RESEARCH TEAM
Liming Chen
Laboratoire d'InfoRmatique en Image et Systèmes d'information (LIRIS)
UMR5205
ECL, Insa, Lyon 2, CNRS
Your technology officer
Yannick Tocquet
Tel. +33(0)4 37 37 42 94