Abstract
Parallel architecture for fast and efficient visual and thermal face recognition fusion biometric system is presented in this paper. Almost all the decision fusion techniques, in the literature, are computationally expensive. Gabor filter technique is used for recognition of features from input image and the database images. As Gabor filter itself has high computational requirements, so we have proposed a parallel solution, which is fast and efficient. Paper also discusses the performance issues of cluster computing based parallel face recognition along with its limitations and benefits.