Dynamic Authentication Technique Using Lip Movements |
Author(s): |
| Patel Hardik , Parul Institute of Engineering & Technology, Limda, Vadodara, India; Maheshwari Nilesh, Parul Institute of Engineering & Technology, Limda, Vadodara, India |
Keywords: |
| Face Recognition, lip ROI detection, Feature extraction, lip reading, syllable tracking, threshold analysis. |
Abstract |
|
Advancement of technology has conquered every strata of human society. In this technological era the protection of important data from being intercepted is growing difficult day by day. Bio-metric security stands to be one of the most secure form of authentication techniques. Such authentication can be seen in the regions with high security such as Defense, Research head-quarters, space-missions etc. Bio-metric authentication today ranges from face recognition to retina recognition. Which are all single stage authentication techniques and cannot be heavily relied upon seeing the current advancement of technology. This concept of multi stage authentication technique which uses a total of three different characteristics / traits for identification of humans not only increases the reliability by many folds but also aid's to the robustness of the system. The first two stages consist of Face recognition and lip pattern recognition, the third stage consists of authentication based on dynamic moments of lips. By tracing and capturing the successive movements of lips during speech, the corresponding words can be detected and can be checked with the data base to authenticate an individual. The captured images are represented as points on a two-dimensional flat manifold that enables us to efficiently define the pronunciation of each word and hence analyze or synthesize the motion of lips. With multiple level of image processing, it becomes possible to set the matching parameters to a very close value and hence it becomes extremely difficult for any other brute force or hacking techniques to break into our system. |
Other Details |
|
Paper ID: IJSRDV1I11050 Published in: Volume : 1, Issue : 11 Publication Date: 01/02/2014 Page(s): 2505-2507 |
Article Preview |
|
|
|
|
