Efficient Real Time Face Tracking Operator Study and Implementation Within Virtex FPGA Technology

In this paper, we present the development of a new real time face-tracking operator. This last is based on motion detection techniques, less complex than techniques based on skin color, which are sensitive to illumination or those based on geometric standards which are efficient but expensive in computation tasks. The developed operator allows a robust face tracking with a considerable reduction in hardware resources, more particularly memory resources, which remains a powerful criterion in the construction of real time embedded systems. Moreover, the operator maps efficiently into a highly pipelined architecture, well suited for an implementation in reconfigurable technology. The developed operator is implemented on a Virtex FPGA architecture. The obtained experimental results show the effectiveness and correctness of our approach.