An efficient parallel approach using OpenCL for pupil detection and pupil size measurement

Pupil detection techniques based on video camera are useful for human monitoring and machine-human interface devices. For example, PD can be used in the field of noninvasive intracranial pressure monitor, fatigue detection etc. During the pupil detection process, image pixels can dramatically influence the detection speed, especially when two cameras are used for acquiring images from both left and right eyes. The low detection speed would impact the application for real-time detections. In this paper, in order to improve the detection speed, a parallel pupil detection and pupil size measurement method using the OpenCL (Open Computing Language) framework for the parallel computing was proposed. The experimental results indicated that the proposed method present excellent efficiency than that of CPU-only platforms without compromising the accuracy of pupil detection and pupil size measurement.

[1]  José Hiroki Saito,et al.  Processing Neocognitron of Face Recognition on High Performance Environment Based on GPU with CUDA Architecture , 2008, 2008 20th International Symposium on Computer Architecture and High Performance Computing.

[2]  Sadaoki Furui,et al.  Harnessing graphics processors for the fast computation of acoustic likelihoods in speech recognition , 2009, Comput. Speech Lang..

[3]  Marium N. Khan,et al.  Noninvasive monitoring intracranial pressure – A review of available modalities , 2017, Surgical neurology international.

[4]  Carlos Hitoshi Morimoto,et al.  Real-time multiple face detection using active illumination , 2000, Proceedings Fourth IEEE International Conference on Automatic Face and Gesture Recognition (Cat. No. PR00580).

[5]  Rolf Adams,et al.  Seeded Region Growing , 1994, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Ryan N. Rakvic,et al.  Parallelizing Iris Recognition , 2009, IEEE Transactions on Information Forensics and Security.

[7]  Ashok A. Ghatol,et al.  Iris recognition: an emerging biometric technology , 2007 .

[8]  Luca Maria Gambardella,et al.  Deep, Big, Simple Neural Nets for Handwritten Digit Recognition , 2010, Neural Computation.

[9]  Takahiro Okabe,et al.  Gaze Estimation From Eye Appearance: A Head Pose-Free Method via Eye Image Synthesis , 2015, IEEE Transactions on Image Processing.

[10]  Zheng Zhang,et al.  FERA 2017 - Addressing Head Pose in the Third Facial Expression Recognition and Analysis Challenge , 2017, 2017 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017).

[11]  Geir Evensen,et al.  The Ensemble Kalman Filter: theoretical formulation and practical implementation , 2003 .

[12]  Qinmu Peng,et al.  Eye Gaze Tracking With a Web Camera in a Desktop Environment , 2015, IEEE Transactions on Human-Machine Systems.

[13]  Keechul Jung,et al.  GPU implementation of neural networks , 2004, Pattern Recognit..

[14]  John Daugman,et al.  New Methods in Iris Recognition , 2007, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[15]  Sungdae Cho,et al.  Design and Performance Evaluation of Image Processing Algorithms on GPUs , 2011, IEEE Transactions on Parallel and Distributed Systems.