Face localization and its implementation on embedded platform

Face localization is a first step for face processing system. A lot of algorithmic work on face processing has already been reported. It has mostly been implemented on computers. For better speed, the algorithms need to be implemented on embedded platforms. The embedded algorithms need to be computationally inexpensive. This paper introduces computationally light face localization, ear and neck separation algorithms for implementation on embedded BeagleBoard-xM platform. Experimental results of the proposed algorithms are presented in the results section along with benchmarking with other contemporary algorithms.