A Video Surveillance system at Security zone

Abouzar Ghasemi, C. N Ravikumar

Abstract


In this paper, we introduce an real time video based face recognition framework at security zone. The created framework recognizes subjects(e.g., travellers/customer) while they are entering a security entryway. Uncontrolled varieties of facial appearance because of occlusion, changing brightening, posture or expression of non cooperative subjects should be handled to correct recognition. To achieve this goal, the framework first detects and tracks the eyes for registration. Then registered subject faces are classified by a local appearance based face recognition algorithm individually. In next step by outcome of confidence scores from each classification are progressively combined to get the identity estimate of the all sequences. We propose three unique measures to weight the contribution of all frames individually to determine the general classification. These are distance to model, distance to second nearest representative, and their composition. We have closed set identification experiments on a database of 125 subjects and the result show that the our framework is capable to get more than 92 percent of correct recognition rate.

Keywords


Video surveillance system, Face detection, Face recognition, Eye tracking, DCT.

Full Text:

PDF

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.