Implementation of Face and Eye Tracking System Using High Level Synthesis

Abstract: This Proposed system is developed by implementing an algorithm proposed by Viola-Jones for face and eye detection and tracking. By monitoring various features of face and eye, attention of person can be estimated. Given system uses Vivado High Level Synthesis (HLS) for implementing Viola-Jones algorithm. Here, Xilinx Zynq-7000 SoC series is used as hardware platform, processing platform. OpenCV libraries which gives optimum real time image processing, is used to obtain required application. Viola-Jones uses Haar cascade classifier which classifies different face inclination of person and system estimates whether a person is attentive or not. This system is running on the ARM Cortex-A9 dual core processor of the Zynq-7000 SoC and with Vivado HLS along with OpenCV libraries, and the results show that the system performs satisfactorily when a face is not inclined. However, if face is inclined at certain angle, system stops detection and tracking of face and eyes.

Keywords: Vivado HLS, Viola-Jones algorithm, OpenCV libraries, Xilinx Zynq-7000 SoC.