The purpose of this research is to provide an evaluation of how facial aging has an effect on image quality. Facial recognition is a growing field in biometrics; it is important to distinguish what characteristics of the face can change without making an impact on the facial recognition system. To determine this, we used the following variables: Eye Separation, Eye Axis Angle, Eye Axis Location, Facial Dynamic Range, and Percent Facial Brightness.
(Spring 2013) Impact of Facial Aging on Image Quality
1. IMPACT OF FACIAL AGING ON IMAGE QUALITY
The purpose of this research is to provide an evaluation of how facial aging has an effect on image quality. Facial recognition is a growing
field in biometrics; it is important to distinguish what characteristics of the face can change without making an impact on the facial
recognition system. To determine this, we used the following variables: Eye Separation, Eye Axis Angle, Eye Axis Location, Facial
Dynamic Range, and Percent Facial Brightness.
Amol Gharte, John Shiver, Peters Drey, Michael Brockly, Stephen Elliott
Overview
Facial Dynamic Range is the ratio of lightest to darkest pixel of the
image in comparison to the average population.
This variable is important in explaining the impact of facial aging on
image quality because it is one of the key attributes that a biometric
device will take into account when making a pass or fail decision on
facial comparisons. The table, Different Variables in Facial Aging,
shows variables such as eye separation, facial brightness, eye axis
location, and eye axis angle are all inconsistent from each time lapse
sample.
This set of images is an example of a time lapse series. The woman had taken a photo
everyday for 5.5 years and the above photos are samples from randomly chosen days.
Video
Eye
Separation
Facial
Brightness
Facial
Dynamic
Range
Eye Axis
Location
Eye Axis
Angle
1 Year I D D D I
8 Months D D N I N
6 Years D I N D D
5.5 Year N N N N N
8 Years N N N N N
Increasing Trend (I), Decreasing Trend (D), No Trend (N)
1 year8 months
6 years 5.5 years
8 years
After taking into account five key variables our studies showed that
facial dynamic range was the most consistent over time. Our
research showed this through the graphs because the slopes of the
facial dynamic range are fairly stable compared to the other
variables. In conclusion, this study has proven that there is no
impact of facial aging on image quality.
Facial Dynamic Range
The graphs are results of studies
from five different individuals
who took time lapse photos
ranging from eight months to
eight years. These are indicators
of how stable facial dynamic
range is in comparison to other
variables.
Video
Incremented Time Between
Picture
Amount of Photos
8 Month Every week 28
8 Years N/A 1470
5.5 Years daily 638
6 Years N/A 880
1 Year daily 366
Different Variables in Facial Aging
Conclusion
Time Lapse Information