隨著老年人口和慢性病病患的增加，以及人們對健康觀念的提升，健康照護（health care）議題變得越來越重要。基於健康照護的重要性，本論文針對老年人於睡眠期間的健康監控和長期臥床病患的壓瘡預防，提出一個應用情境分析技術於視訊影像之健康照護系統，此系統含即時且非接觸式之睡眠者呼吸量測系統與臥床病患翻身偵測系統。

呼吸是評估個人健康與否的一個重要指標。然而，現存之呼吸量測技術，其設備皆須接觸使用者以擷取呼吸資訊。這不僅會引起使用者的不舒適感，更會影響其睡眠品質。本論文提出一個非接觸式呼吸量測系統，用以即時測量使用者的呼吸資訊。此系統以偵測呼吸所引起的胸腹部之週期性起伏變化為根據，並依此變化擷取呼吸資訊。由於呼吸造成的起伏變化非常細微，因此很容易受到身體移動所干擾。同時，身體的移動亦可能造成影像中胸腹部位置與臥姿的改變。為了精確的量測呼吸，系統必須能夠動態的定位出影像中胸腹部的區域。再者，不同的臥姿會導致影像中呼吸方向的不同，所以系統還必須能夠決定呼吸的方向。為了讓系統能夠正確的量測呼吸，我們提出三個子系統。（1）身體移動偵測子系統：用以辨別使用者的呼吸移動與身體移動（含翻動）。為了精確地切割出前景和偵測身體移動，我們提出一個以情境為基礎的背景相減（context-based background subtraction）機制。此機制利用一個可適應性的時間常數方法（adaptive time constant scheme）以正確地辨識睡眠者身體的移動狀態（移動（moving）或靜止（motionless）），與快速地偵測出兩個狀態間的狀態轉換。同時，為了從擷取的移動資訊中動態地決定工作參數，我們也提出一個使用有限狀態以控制隱藏碼可夫模型分類結果（finite state controlled hidden markov model）的機制。（2）呼吸情境分析子系統:其目的為動態地決定和定位出影像中睡眠者的呼吸方向和呼吸區域。首先，我們使用移動偵測常用之光學流估測（optical flow estimation）方法，得到兩張睡眠者呼吸移動量差異最大的影像，並利用平均位移演算法（mean shift algorithm）定位出呼吸區域。（3）以移動向量為基礎的即時呼吸量測子系統：藉此加速系統計算和分析睡眠者呼吸資訊之效能。本子系統會偵測呼吸區域中的特徵熱區（feature zones），並提出一個改良式之快速光學流（improved fast optical flow）計算方法得到呼吸曲線，並藉此分析睡眠者之呼吸次數、深度與兩兩呼吸的時間間隔。

對於臥床病患翻身偵測系統，其目的為偵測長期臥床病患是否定期翻身以避免壓瘡的發生。對一個臥床者言，當他翻動身體時，此翻動必然造成身體的移動(movement)；然而，當臥床者移動他的身體時，則此移動不一定是身體的翻動。為了進一步確認，我們利用一般人其胸部之正面寬度大於側面寬度的生理結構特徵，用以判斷移動前與移動後之臥姿是否改變。為了正確的偵測臥床者的翻身，系統必須克服許多挑戰。由於臥床者可能隨時且隨意地移動他的身體，這樣的移動通常會改變臥床者的胸部位置。因此系統要能夠動態地定位出影像中臥床者新的胸部區域。再者，如何從二維影像精確地計算出胸部的寬度資訊。最後，移動前與移動後的胸部寬度差異性只能用來判斷移動前後的正躺與側躺之轉變，並無法得知側面為側左或側右。為了解決這些問題，我們提出一個翻身偵測系統，以偵測九種翻身行為，此系統包含三個子系統。(1) 身體移動偵測子系統：用來立即地偵測臥床病患是否移動他的身體，而且辨識病患目前的移動狀態（移動或靜止）。在此，我們使用呼吸量測系統之身體移動情境偵測技術來達到這個目的。（2）上半身特徵擷取子系統：用來獲得上半身之寬度資訊以判別臥床病患為正躺或側躺。我們首先使用頭部定位和追蹤技術定位出頭部位置和大小，並藉此找出影像中病患上半身長度之區域。再者，我們使用陰影造型 (shape-from-shading)技術重建出病患的立體身體結構圖。最後由上半身長度之區域以及身體結構圖，切割出精確的上半身熱區(torso zone)，並計算此熱區的寬度當作上半身的特徵。(3)翻身行為辨識子系統：目的為偵測九種翻身行為，並決定偵測到的翻身行為是否滿足壓瘡的預防法則。首先，我們利用移動前與移動後的上半身寬度差異判別正躺與側躺的轉換。再者，使用光學流估測（optical flow estimation）方法以獲得身體的翻身方向，最後正臉偵測（frontal face detection）技術被用來協助系統偵測臥床病患的翻身行為。

由於此健康照護系統使用非接觸式技術與近紅外光源和攝影機，因此不會限制使用者的活動和造成不舒適感，也不會打擾其睡眠和無法在昏暗的環境工作。實驗結果證明，本論文所提出的健康照護系統確實能夠精確的量測臥床睡眠者的呼吸資訊，以及辨識臥床者的身體翻身行為，藉此達到健康照護的目的。

With the increase of the global elderly population and the number of patients with chronic disease, and the promotion of health concept of people, healthcare is gaining more and more attention. Based on the importance of healthcare, the paper focuses on the health monitoring of sleeping elderly and the pressure-ulcer prevention of bedridden patient to propose a video-based context analysis for health care system. The system consists of the contact-free and real-time sleeper’s respiration measurement system and lying patient’s body-turning detection system.

Respiration is a critical data for estimating a person’s health. However, the existed respiring measurement methods must touch user to sense the respiration. It not only makes the user uncomfortable but also disturbs his/her sleep. This paper proposes a contact-free respiration measurement system to measure the sleeper’s respiration on bed. The system is based on the periodically raising and falling motion of the user’s chest or abdomen caused by respiration. To make the system work well, some problems must be considered. As well known, the movement caused by respiration is only tiny. And, the sleeper may move and turn their bodies, consciously or unconsciously, during sleep. These movements involve large motions compared with the respiring movement, and they also usually cause the sleeper to change body position and posture. Therefore, it is necessary to dynamically locate the new position of the respiration region and decide the representative respiratory motion direction for analysis. In addition, the system must have ability to real-time measure the respiration. To resolve these problems, we propose a respiration-measurement system, including three subsystems. (1) The body-motion-context detection subsystem distinguishes between respiratory movement and non-respiratory body movement. To optimally segment the foreground and detect body motion, we propose a context-based background subtraction with an adaptive time constant scheme. This scheme can accurately recognize the status of the sleeper’s body movement—moving or motionless—and quickly detect the transition from one status to the other. To dynamically determine the working parameters based on the extracted motion information, we develop a finite state controlled hidden markov model. (2) The goal of the respiration-context detection subsystem is to dynamically detect the representative direction of the sleeper's respiration motion and locate the respiratory region in image. This subsystem searches the pair of representative frames, which are the two frames with the maximum motion difference during the sleeper's respiration, derived using the optical flow estimation method. The mean shift algorithm is then used to locate the respiratory region. (3) The motion-vector-based real-time respiration measurement subsystem speeds up the system’s performance for the subsequent respiration analysis. The subsystem detects feature zones in the respiratory region and then uses an improved fast optical flow method to obtain the respiratory curve to analyze the respiration frequency, depth, and the time interval between two respirations of patient.

The lying patient’s body-turning detection system is proposed to detect whether the lying posture of patient on bed has changed to prevent the pressure ulcer. For a lying patient, when he turns the body, his body must be moved; however, when he moves the body, this movement may not cause a body-turning. To check it further, we utilize a characteristic of physical structure, the torso's width being bigger from the front than from the side, to identify whether the lying posture before and after movement changes. To detect the body-turning behavior of lying patient, the system must face some challenges. Because a lying patient may spontaneously move his body at any time, this movement may change the torso’s position. Therefore, it is necessary to dynamically locate the new position of torso in image. Besides, it is difficult to compute the exact width of torso from the 2-dimension image. Lastly, the difference of torso’s width before and after movement is only used to indentify the change of lying posture between supine and lateral, which is unable to distinguish that the lateral is lateral-left or lateral-right. To resolve these problems, we propose a body-turning detection system to detect 9 body-turning behaviors exactly. This system includes three subsystems: (1) the body motion detection subsystem which is used to immediately detect whether the lying patient moves the body and indentify the patient’s current movement status (moving or motionless). Here, we use the body-motion-context detection technique in the respiration measurement system to achieve it. (2) The torso’s feature extraction subsystem which is proposed to obtain the torso’s width to recognize that a lying patient is supine or lateral. It uses the head location and tracking techniques to locate the head’s position and size to further decide the torso-length region of the patient in image. Then, the shape-from-shading method is used to reconstruct the stereo body structure map of the patient. Finally, the torso-length region and the body structure map are used to segment the exact torso zone, and then the width of this zone is computed as the representative feature of torso. (3) The body-turning behavior recognition subsystem which is developed to detect 9 body-turning behaviors and decide whether the detected body-turning behavior satisfies the preventing rule of pressure ulcer. Firstly, the width difference of the torso before and after the movement is used to recognize the switch of supine and lateral postures. The optical flow estimation method is then adopted to detect the body-turning direction, and finally the frontal face detection mechanism is applied to help the system detect the body-turning behavior.

Because the proposed healthcare system uses the contact-free technique and a near-infrared camera and lighting, it neither restricts the users’ movement and causes them discomfort nor disturbs their sleep and stop work in dim condition. Experimental results show that the methods of the proposed video-based context analysis for health care system can achieve satisfactory performance for measuring the respiratory information of sleeper on bed and recognizing the body-turning behavior of long-term lying patient.

[1] J. Maklebust and M. Y. Sicggreen, Pressure ulcers: Guidelines for prevention and management (3rd ed.), 2001.
[2] G. B. Moody, R. G. Mark, A. Zoccola and S. Mantero, “Derivation of Respiratory Signals from Multi-lead ECGs,” Computers in Cardiology, Vol. 12, pp. 113-116, 1985.
[3] L. Scalise, I. Ercoli, and P. Marchionni, “Optical method for measurement of respiration rate,” Proceedings of the IEEE International Workshop on Medical Measurements and Applications, pp. 19-22, 2010.
[4] J. Park, J. H. Chang, S. Ahn, Y. K. Pak, S. Han, and J. J. Pak, “Array type dissolved oxygen sensor and measurement system for simultaneous measurement of cellular respiration level,” International Conference on Solid-State Sensors, Actuators and Microsystems (Transducers), pp. 1015-1018, 2009.
[5] E. P. Scilingo, A. Gemignani, R. Paradiso, N. Taccini, B. Ghelarducci, and D. D. Rossi, “Performance Evaluation of Sensing Fabrics for Monitoring Physiological and Biomechanical Variables,” IEEE Transactions on Information Technology in Biomedicine, Vol. 9, No. 3, pp. 345-352, 2005.
[6] D. Barschdorff and W. Zhang, “Respiratory Rhythm Detection with Photoplethysmographic Methods,” Engineering in Medicine and Biology Society. Engineering Advances: New Opportunities for Biomedical Engineers. Proceedings of the Annual International Conference of the IEEE, Vol. 2, pp. 912-913, 1994.
[7] C. R. Merritt, H. T. Nagle, and E. Grant, “Textile-Based Capacitive Sensors for Respiration Monitoring,” IEEE Sensors Journal, Vol. 9, No. 1, pp. 71-78, 2009.
[8] M. Kermit, A. J. Eide, T. Lindblad, and K. Waldemark, “Treatment of obstructive sleep apnea syndrome by monitoring patient airflow signals,” Pattern Recognition Letters, Vol. 21, pp. 277-281, 2000.
[9] B. A. Phillips, M. I. Anstead, and D. J. Gottlieb, “Monitoring sleep and breathing: Methodology. Part I: Monitoring breathing,” Clinics in Chest Medicine, Vol. 19, No. 1, pp. 203-212, 1998.
[10] J. R. Stradling, “New approaches to monitoring sleep-related breathing disorders,” Sleep, Vol. 18, No. 3, pp. 77-84, 1996.
[11] B. H. Taha, J. A. Dempsey, S. M. Weber, M. S. Badr, J. B. Skatrud, T. B. Young, A. J. Jacques, K. C. Seow, “Automated detection and classification of sleep-disordered breathing from conventional polysomnography data,” Sleep, Vol. 20, No. 11, pp. 991-1001, 1997.
[12] A. Bartolo, B. D. Clymer, J. A. Golish, and R. C. Burgess, “The polysomnogram assay: A method to represent the overnight polysomnogram in a condensed fromat,” Computers and Biomedical Research, Vol. 33, No. 2, pp. 110-125, 2000.
[13] R. K. Millard, “Inductive plethysmography components analysis and improved noninvasive postoperative apnoea monitoring,” Physiological Measurement, Vol. 20, No. 2, pp. 175-186, 1999.
[14] P. Varady, L. Nagy, L. Szilagyi, “On-line detection of sleep apnea during critical care monitoring,” Engineering in Medicine and Biology Society. Proceedings of the Annual International Conference of the IEEE, Vol.2, pp.1299-1301, 2000.
[15] P. Várady, T. Micsik, S. Benedek, and Z. Benyó, “A novel method for the detection of apnea and hypopnea events in respiration signals,” IEEE Transactions On Biomedical Engineering, Vol. 49, No. 9, pp. 936-942, 2002.
[16] K. Nepal, E. Biegeleisen and T. Ning, “Apnea detection and respiration rate estimation through parametricmodelling,” Proceedings of the IEEE Annual Northeast Bioengineering Conference, pp. 277-278, 2002.
[17] M. I. Restrepo, S. Bhandari, and T. Ning, “Classification of Respiration Episodes using Fuzzy Logic,” Bioengineering Conference. Proceedings of the IEEE Annual Northeast, pp. 133-134, 2006.
[18] H. Kametani, M. Takao and T. Setoguchi, “Characteristic evaluation of the flowmeter using the Wells turbine for measurement of respiration,” Proceedings of the SICE Annual Conference, Vol. 1, pp. 531-534, 2002.
[19] Y. Wang, H. Baghaei, H. Li, Y. Liu, T. Xing, J. Uribe, R. Ramirez, S. Xie, S. Kim, and W. H. Wong, “A Simple Respiration Gating Technique and Its Application in High-Resolution PET Camera,” IEEE Transactions on Nuclear Science, Vol. 52, No. 1, pp. 125-129, 2005.
[20] B. W. Carlson, V. J. Neelon and H. Hsiao, “Evaluation of a non-invasive respiratory monitoring system for sleeping subjects,” Physiological Measurement, Vol. 20, pp. 53-63, 1999.
[21] W. Chen, X. Zhu, T. Nemoto, Y. Kanemitsu, K. Kitamura and K. Yamakoshi, “Unconstrained detection of respiration rhythm and pulse rate with one under-pillow sensor during sleep,” Medical and Biological Engineering and Computing, Vol. 43, pp. 306-312, 2005.
[22] T. Usui, A. Matsubara, and S. Tanaka, “Unconstrained and noninvasive measurement of heartbeat and respiration using an acoustic sensor enclosed in an air pillow,” SICE Annual Conference, Vol. 3, pp. 2648-2651, 2004.
[23] S. Brady, L. E Dunne, R. Tynan, D. Diamond, B. Smyth, G. M. P. O'Hare, “Garment-based monitoring of respiration rate using a foam pressure sensor,” Proceedings of the IEEE International Symposium on Wearable Computers, pp. 214-215, 2005.
[24] I. Svetlana, H. Y. Mammo, W. A. John, E. H. Michael, H. M. Carl, and R. P. Scott, “A gated deep inspiration breath-hold radiation therapy technique using a linear position transducer,” Journal of Applied Clinical Medical Physics, Vol. 6, pp. 61-70, 2005.
[25] V. M. Lubecke and Olga Borić-Lubecke, “Wireless Technologies in Sleep Monitoring,” IEEE Radio and Wireless Symposium, pp. 135-138, 2009.
[26] M. A. Gennert, J. K. Ho, A. C. Quina, J. H. Wang, P. P. Bruyant, and M. A. King, “Feasibility of tracking patient respiration during cardiac SPECT imaging using stereo optical cameras,” IEEE Nuclear Science Symposium Conference Record, Vol. 5, pp. 3170-3172, 2003.
[27] R. Paradiso, G. Loriga, and N. Taccini, “A Wearable Health Care System Based on Knitted Integrated Sensors,” IEEE Transactions on Information Technology in Biomedicine, Vol. 9, No. 3, pp. 337-344, 2005.
[28] J. C. Lin, “Noninvasive microwave measurement of respiration,” Proceedings of the IEEE, Vol.63, No.10, pp. 1530-1530, Oct. 1975.
[29] Y. Huang, K. M. Chen, J. Zhang, and Y. Dai, “Fast accurate measurement of phase and amplitude using EM microprocessor-based zero-balance system,” International Symposium Antennas and EM Theory, Xi'an, China, Aug. 19-22, 1997.
[30] Y. Xiao, J. Lin, O. Boric-Lubecke, and V. M. Lubecke, “A Ka-Band Low Power Doppler Radar System for Remote Detection of Cardiopulmonary Motion,” Proceedings of the IEEE Engineering in Medicine and Biology Annual Conference, pp. 7151-7154, Sept. 1-4, 2005.
[31] N. D. Duffy, S. McGowan, and G. B. Drummond, “Optical non-contact measurement of abdominal volume change during anaesthesia,” Optical Techniques and Biomedical Applications, IEE Colloquium on, pp. 8/1-8/4, 14 Jun 1991.
[32] Z. Zhu, J. Fei, and I. Pavlidis, “Tracking human breath in infrared imaging,” Proceedings of the IEEE Symposium on Bioinformatics and Bioengineering, pp. 227-231, 2005.
[33] R. Murthy and I. Pavlidis, “Noncontact measurement of breathing function,” IEEE Engineering in Medicine and Biology Magazine, Vol. 25, Issue 3, pp. 57-67, 2006.
[34] D. U. Silverthorn, “Respiratory physiology,” Human Physiology: An Integrated Approach, 2nd ed. Englewood Cliffs, NJ: Prentice-Hall, pp. 498-508, 2001.
[35] B. K. P. Horn and B. G. Schunck, “Determining optical flow,” Artificial Intelligence, Vol. 17, pp. 185-203, 1981.
[36] K. Fukunaga and L. D. Hostetler, “The estimation of the gradient of a density function, with applications in pattern recognition,” IEEE Transactions on Information Theory, Vol. 21, pp. 3240, 1975.
[37] A. J. Lipton, H. Fujiyoshi, and R. S. Patil, “Moving target classification and tracking from real-time video,” Proceedings of IEEE Workshop Applications of Computer Vision, pp. 8-14, 1998.
[38] N. Friedman and S. Russell, “Image segmentation in video sequences: a probabilistic approach,” Proceedings of Conference on Uncertainty in Artificial Intelligence, pp. 175-181, 1997.
[39] B. P. L. Lo and S. A. Velastin, “Automatic congestion detection system for underground platforms,” Proceedings of ISIMP, pp. 158-161, 2001.
[40] A. Elgammal, R. Duraiswami, D. Harwood, and L. S. Davis, “Background and foreground modeling using nonparametric kernel density estimation for visual surveillance,” Proceedings of the IEEE, Vol. 90, No. 7, pp. 1151-1163, 2002.
[41] C. R. Wren, A. Azarbayejani, T. Darrell, and A. P. Pentland, “Pfinder: Real-time tracking of the human body,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 19, No. 7, pp. 780-785, 1997.
[42] C. Stauffer, and W. E. L. Grimson, “Adaptive background mixture models for real-time tracking,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Vol. 2, pp. 246-252, 1999.
[43] L. E. Baum and T. Petrie, “Statistical inference for probabilistic functions of finite state Markov chains,” Annals of Mathematical Statistics, Vol. 37, pp. 1554-1563, 1966.
[44] L. R. Rabiner, “A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition,” Proceedings of the IEEE, Vol. 77, No. 2, pp. 257-286, 1989.
[45] C. W. Oomens, O. F. J. T. Bressers, E. M. H. Bosboom, C. V. C. Bouten, and D. L. Bader, “Can loaded interface characteristics influence strain distribution in muscle adjacent to bony prominences,” Computer Methods in Biomechanics and Biomedical Engineering, Vol. 6, No. 3, pp. 171-180, 2003.
[46] National Pressure Ulcer Advisory Panel, “Pressure ulcers prevalence, cost and risk assessment,” Decubitus, Vol. 2, No. 2, pp. 24-28, 1989.
[47] C. W. Oomens, R. G. Breuls, C. V. C. Bouten, F. P. T. Baaijens, and D. L. Bader, “A hierarchical approach to study damage evolution in skeletal muscle - a contribution to pressure ulcer research,” Summer Bioengineering Conference, 2003.
[48] T. J. Phillips, “Chronic cutaneous ulcers: etiology and epidemiology,” Journal of Investigative Dermatology, Vol. 102, No. 6, 38s-41s, 1994.
[49] P. W. Smith, J. M. Black and S. B. Black, “Infected pressure ulcers in the long-term-care facility,” Infection Control and Hospital Epidemiology, Vol. 20, No. 5, pp. 358-363, 1999.
[50] J. Maklebust and M. Y. Sicggreen, Pressure ulcers: Guidelines for prevention and management (3rd ed.), 2001.
[51] R. S. Shutes, “Bed Mounted Device for Patient Positioner,” U.S. Patent Number 5524304, 1996.
[52] Fuller, “Bed Patient Turning Lifting and Transporing Apparatus with Mobile Folding and Knockdown Fram,” U.S. Patent Number 5544371, 1996.
[53] Marmor, “Apparatus for Turning a Patient in Bed,” U.S. Patent Number 5673443, 1997.
[54] Behr, “Patient Positioning Device,” U.S. Patent Number 6065162, 2000.
[55] Takaura, “Method for Changing the Posture of a Patient on a Nursing Bed,” U.S. Patent Number 6295666 B1, 2001.
[56] T. W. Votel, “Patient Transfer and Repositioning System,” U.S. Patent Number 6341393 B1, 2002.
[57] B. W Wheeler, “Patient Rotation System”, U.S. Patent Number 6393636, 2002.
[58] C. H. Wei, “The Prototype Making of Patient Bed with Functions of Automatic Turning and Massaging Bedridden People," HUhttp://ir.lib.stut.edu.tw/handle/987654321/2816UH, 2002.
[59] B. Roy, A. Basmajian, and H. H. Asada, “Repositioning of a rigid body with a flexible sheet and its application to an automated rehabilitation bed,” IEEE Transactions on Automation Science and Engineering, Vol. 2, No. 3, pp. 300-307, Jul. 2005.
[60] A. Basmajian, E. E. Blanco, and H. H. Asada, “The marionette bed: Automated rolling and repositioning of bedridden patients,” Proceeding IEEE International Conference on Robotics and Automation, Vol. 2, pp. 1422-1427, 2002.
[61] S. W. Peng, F. L. Lian, and L. C. Fu, “Development of Multi-Functional Robotic Test-Bed for Post-Surgical Healthcare Room,” IEEE International Conference on Rehabilitation Robotics, 2009.
[62] C. h. Yi, H. S. Kim, W. G. Yoo, M. H. Kim, and O. Y. Kwon, “The Effects of Different Types of Automated Inclining Bed and Tilt Angle on Body-Pressure Redistribution,” Advances in Skin & Wound Care, Vol. 22, Issue 6, pp. 259-264, 2009.
[63] B. K. P. Horn, “Shape from Shading: A Method for Obtaining the Shape of a Smooth Opaque Object from One View,” Ph. D. Thesis, Massachusetts Institute of Technology, 1970.
[64] M. J. Brooks and B. K. P. Horn, “Shape and source from shading,” Proceedings of International Joint Conference on Artificial Intelligence, pp. 932-936, 1985.
[65] B. K. P. Horn, R. S. Szeliski, and A. L. Yuille, “Impossible shaded images,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 15, No. 2, pp. 166-170, 1993.
[66] F. Moreno, A. Tarrida, J. Andrade-Cetto, and A. Sanfeliu, “3D real-time head tracking fusing color histograms and stereovision,” International Conference on Pattern Recognition, Vol. 1, pp. 368-371, 2002.
[67] P. Fieguth, and D. Terzopoulos, “Color-based tracking of heads and other mobile objects at video frame rates,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 21-27, 1997.
[68] M. Bohme, M. Haker, T. Martinetz, and E. Barth, “Head tracking with combined face and nose detection,” International Symposium on Signals, Circuits and Systems, pp. 1-4, 2009.
[69] S. Birchfield, “An Elliptical Head Tracker,” Proceedings of the Asilomar Conference on Signals, Systems, and Computers, pp. 1710-1714, 1997.
[70] N. Otsu, "A Threshold Selection Method from Gray-Level Histograms," IEEE Transactions on Systems, Man, and Cybernetics, Vol. 9, No. 1, pp. 62-66, 1979.