Testanlagen
Roboter Test StreckeRoboter Test StreckeRoboter Test StreckeUnterwasser TestbedSpace Testbed
Projektkoordination
Talkarchive

Vortragsdetails

Ort:DFKI Bremen
Robotics Innovation Center
Robert-Hooke-Str. 5
Konferenzraum 117
Opens internal link in current windowAnfahrtsbeschreibung
Federico M. Sukno, PhD,  Networking Center on Biomedical Research - CIBER-BBN
Automatic assessment of eye-blinking patters for drivers inattention through statistical shape models
(Abstract)
Several studies have related the alertness of
individuals to their eye-blinking patterns. Accurate and automatic
quantification of eye-blinks can be of much use in monitoring people
at jobs that require high degree of alertness, such as that of a
driver of a vehicle. However, systems to monitor eye blinking are
usually built using infrared devices which, apart from the need for
specialized and expensive equipment, make the setup not
user-friendly and provides substantial visual obstruction to the
individual. This talk analyzes the use of a non-intrusive system
based on facial biometrics techniques to accurately detect and
quantify eye-blinks. A method based on statistical shape models is
presented, discussing current results and the main challenges ahead,
together with possible solutions. Given a video sequence from a
standard camera, the proposed procedure can output blink frequencies
and durations, as well as the PERCLOS metric (the percentage of time
the eyes are at least 80 percent closed), currently considered the most
reliable and valid determination of a drivers alertness level.