Project title: Periocular Biometrics Data Analytics
2D Camera - Canon EOS 5D Mark III Kit
3D White Light Scanner – Artec Eva Lite
Periocular biometrics is receiving enormous attention due to its benefits of reduced storage space, reliable acquisition, increased speed and reduced computation cost. Currently an exclusive database consisting of periocular region images in not publicly available. Research activity related to periocular region based authentication systems is carried out using the available face or iris image database. Extraction of periocular region from face image requires laborious preprocessing and segmentation procedures. In contrast, retrieval of periocular information using iris imaged datasets give subtle or little periocular regions. Algorithms developed using different databases (imaged under different environments) are independently evaluated by each researcher. There is a need to standardize the performance comparison of various periocular algorithms under different unconstrained environments. The development of a periocular database promotes periocular region investigation using robust authentication/recognition algorithms. The project aims at collecting/constructing 2D and 3D periocular images from large number of subjects under constrained/unconstrained environment and be made publicly available.
- To promote research in the newly emerging technology, periocular region based biometric system.
- To construct database exclusively with periocular images simulating unconstrained environment with variations in illumination, pose, expression, distance and introduce occlusions.
- To evaluate the performance of the developed periocular database using standard algorithms.
- To make the database publicly available and provide support/maintenance.
- Periocular region based database for biometric applications (Construction and maintenance)
- Journal publication with good impact factor (At least one)
Grants provided by TEQIP-II worth of Rs. 12 lakhs.