Global face Recognition Strategy based on multiple recognition systems combination
Fahd Idrissi Khamlichi, Rajae Aboulaïch · International journal of imaging and robotics · 2011
In the last decade the high security has become a major problem for government facilities (military, nuclear, air port, banks..), therefore, face recognition has become a popular area of research in computer vision. A general statement of the face recognition problem (in computer vision) can be formulated as follows: Given still or video images of a scene, identify or verify one or more persons in the scene using a stored database of faces. The face recognition systems are generally used for security and for real time systems, so performing those recognition systems and optimizing existing results are always sought. In literature, several holistic, hybrid and feature-based methods of face recognition are widely found such as PCA, LDA, SVM, LVQ,…etc, the performance of those methods depends on many parameters. This paper presents a new approach for face recognition based on multiple existing recognition systems combination, in essence, this strategy performs the two important keys of real time recognition, by: • Optimizing the recognition time. • Performing the recognition. The main contribution of this strategy is seen in the fact that it meets the needs of real time recognition and makes the recognition more improved.