Optical flow image analysis of facial expressions of human emotion: forensic applications

Carmen J. Duthoit; Tamara Sztynda; Sara K.L. Lal; Budi T. Jap; J.I. Agbinya

1st International ICST Conference on Forensic Applications and Techniques in Telecommunications, Information and Multimedia

Research Article

Optical flow image analysis of facial expressions of human emotion: forensic applications

Cite: BibTeX Plain Text

@INPROCEEDINGS{10.4108/e-forensics.2008.2743,
    author={Carmen J. Duthoit and Tamara Sztynda and Sara K.L. Lal and Budi T. Jap and J.I. Agbinya},
    title={Optical flow image analysis of facial expressions of human emotion: forensic applications},
    proceedings={1st International ICST Conference on Forensic Applications and Techniques in Telecommunications, Information and Multimedia},
    publisher={ACM},
    proceedings_a={E-FORENSICS},
    year={2010},
    month={5},
    keywords={Emotion facial expression forensics surveillance image analysis optical flow physiological measurement.},
    doi={10.4108/e-forensics.2008.2743}
}

Carmen J. Duthoit
Tamara Sztynda
Sara K.L. Lal
Budi T. Jap
J.I. Agbinya
Year: 2010
Optical flow image analysis of facial expressions of human emotion: forensic applications
E-FORENSICS
ACM
DOI: 10.4108/e-forensics.2008.2743

Carmen J. Duthoit¹^,*, Tamara Sztynda²^,*, Sara K.L. Lal²^,*, Budi T. Jap², J.I. Agbinya³^,*

1: Dept. Medical and Molecular Biosciences (MMB), University of Technology, Sydney. P.O. Box 123 Broadway NSW 2007 +61438317670
2: MMB, Centres of IMS, Engineering and Forensic Science University of Technology, Sydney P.O. Box 123 Broadway NSW 2007 +6129514 4157, 1592, 8291
3: Centre for Real-Time Information Networks (CRIN), Engineering University of Technology, Sydney P.O. Box 123 Broadway NSW 2007 +61295142495

*Contact email: Carmen.J.Duthoit@uts.edu.au, Tamara.Sztynda@uts.edu.au, Sara.Lal@uts.edu.au, agbinya@eng.uts.edu.au

Abstract

The objective of this study was to induce emotions in individuals to determine if specific facial movements could be detected and analysed by the optical flow technique. This analysis is in the form of motion vector plots. The procedure ascertains if specific emotions can be defined as a set of facial movements which are common to most people when they experience a particular emotion. 'Emotion vector maps' would then be established for specific emotions. These vector sets could then be applied to automated facial image analysis for security/forensic purposes.

Individuals were videotaped while watching emotion-inducing short films. After the viewing of each short film, volunteers completed a brief self-reporting questionnaire to establish the emotions they experienced. The image sequences were then analysed according to emotion, by using optical flow analysis. Results were statistically analysed.

Trends and analyses are presented in relation to security and video surveillance. Issues and the development of pattern recognition systems applied to human facial images for security purposes are briefly discussed.

Keywords: Emotion facial expression forensics surveillance image analysis optical flow physiological measurement.

Published: 2010-05-16
Publisher: ACM
Modified: 2010-05-16

: http://dx.doi.org/10.4108/e-forensics.2008.2743