A Robust Non-cooperative Face/Iris Recognition Scheme from a Distance in Maritime Domain

Period of Performance: 07/19/2007 - 05/19/2008


Phase 1 STTR

Recipient Firm

Intelligent Automation, Inc.
15400 Calhoun Dr, Suite 190
Rockville, MD 20855
Firm POC
Principal Investigator

Research Institution

Rensselaer Polytechnic Institute
110 8th Street
Troy, NY 12180
Institution POC


Intelligent Automation Inc. (IAI), its academic partner, Rensselaer Polytechnic Institute with Dr. Qiang Ji as the lead and its industrial partner Sarnoff Corporation with Dr. James R. Matey as the lead, propose a robust, automatic, innovative scheme for face recognition from a distance of as far as 50m and iris recognition from a distance of upto 10m to be used in maritime domain for recognition of non-cooperative individuals in real-time. The image sensors of the proposed face and iris recognition scheme consist of a dual camera system. One of the cameras is a low-resolution CCD camera. The image obtained from this camera is used for detecting and continuously multitracking potential targets within the wide field of view (WFOV). The second camera provides narrow field of view (NFOV) image (CCD or NIR camera). Different from the WFOV camera, this camera provides a high-resolution image enabling eye detection (as far as 50m) and iris recognition from distances up to 10m. In the proposed scheme, a multimodality fusion approach is also proposed for fusing the face and iris biometrics. The proposed face recognition scheme can be made functional for day and night operations (24/7) by replacing the CCD cameras with high-resolution long-distance infrared cameras.