Implementation and Optimization of the Eye Gaze Tracking System Based on DM642

Real-time eye detection is important for many HCI applications, including eye gaze tracking, auto stereoscopic displays, video conferencing, face detection, and recognition. Currently, the eye gaze tracking systems are usually implemented on general purpose processors. As eye gaze tracking algorithms move from research labs to the real world, power consumption and cost become critical issues. This motivates searching for implementations using a digital signal processor (DSP). Our goal in this paper is to explore the feasibility of implementing DSP-based eye gaze tracking systems. To achieve this goal, we implement a fully automatic eye gaze tracking system on Texas Instruments´ TMS320DM642 DSP, profile performance, and analyze opportunities for optimization. Our experiments demonstrate that well-optimized eye gaze tracking implementations on DSP can be an effective choice for embedded eye gaze tracking products.