Smart Vision & and Sensing
Department of System Cybernetics, Graduate School of Advanced Science and Engineering
GPU-Based High-Frame-Rate Face Tracking
This study propose a high-speed vision system that can be applied to real-time face tracking at 500 fps using GPU acceleration of a boosting-based face tracking algorithm. By assuming a small image displacement between frames, which is a property of high-frame-rate (HFR) vision, we develop an improved boosting-based face tracking algorithm for fast face tracking by enhancing the Viola-Jones face detector. In the improved algorithm, the size and position of a face pattern to be tracked in an image can be efficiently extracted by reducing the number of window searches for Haar-like features and by combining skin color extraction with Haar classifiers. The improved boosting-based face tracking algorithm is implemented on a GPU-based high-speed vision platform, and face tracking can be executed in real time at 500 fps for an 8-bit color image of 512×512 pixels. In order to verify the effectiveness of the developed face tracking system, we install it on a two-axis mechanical active vision system and perform several experiments for tracking face patterns.
Human face tracking using a 2-DOF active vision
(MPEG movie (9.6M))
- Idaku Ishii, Tomoki Ichida, Qingyi Gu, and Takeshi Takaki : 500-fps Face Tracking System, Journal of Real-Time Image Processing, doi: 10.1007/s11554-012-0255-8 (online first) (2012)
- Idaku Ishii, Hiroki Ichida, and Takeshi Takaki : GPU-based Face Tracking at 500 fps, Proc. IEEE Int. Conf. on Image Processing, pp.565-568, 2011.