Visual Analysis of Faces with Application in Biometrics, Forensics and Health Informatics

Computer vision-based analysis of human facial video provides information regarding to expression, diseases symptoms, and physiological parameters such as heartbeat rate, blood pressure and respiratory rate. It also provides a convenient source of heartbeat signal to be used in biometrics and forensics. This thesis is a collection of works done in five themes in the realm of computer vision-based facial image analysis: Monitoring elderly patients at private homes, Face quality assessment, Measurement of physiological parameters, Contact-free heartbeat biometrics, and Decision support system for healthcare. The work related to monitoring elderly patients at private homes includes a detailed survey and review of the monitoring technologies relevant to older patients living at home by discussing previous reviews and relevant taxonomies, different scenarios for home monitoring solutions for older patients, sensing and data acquisition techniques, data processing and analysis techniques, available datasets for research and development, and current challenges and future research directions. Face quality assessment theme include works related to the application of a face quality assessment technique in acquiring high quality face sequence in real-time and alignment of face for further analysis. In measuring physiological parameters, two parameters are considered among many different physiological parameters: heartbeat rate and physical fatigue. Though heartbeat rate estimation from video is available in the literature, this thesis proposes an improved method by using a new heartbeat footprint tracking approach in the face. The thesis also introduces a novel way of analyzing heartbeat traces in facial video to provide visible heartbeat peaks in the signal. A method for physical fatigue time offset detection from facial video is also introduced. One of the major contributions of the thesis is introducing heartbeat signal from facial video as a novel biometric trait. The way to extract and utilize this biometric trait in person recognition and face spoofing detection is described. In the last part, the thesis introduces an approach for generating facial expression log as a decision support tool by employing a face quality assessment technique to reduce erroneous expression rating. Despite of the solutions introduced in this thesis, ample of new research questions have brought forward to be solved in advancing the areas of health informatics, biometrics and forensics.