Loading Events

« All Events

  • This event has passed.

Fall 2011 GRASP Seminar – David Brainard, University of Pennsylvania, “The Human Demosaicing Algorithm”

October 7, 2011 @ 11:00 am - 12:00 pm

Abstract: The human visual system shares with most digital cameras the design
feature that color information is acquired via spatially interleaved
sensors with different spectral properties.  That is, the human retina
contains three distinct spectral classes of cone photoreceptors, the
L-, M-, and S-cones, and cones of these three classes are spatially
interleaved in the retina.  Similarly, most digital cameras employ a
design with interleaved red, green, and blue sensors.  In each case,
generating a full color image requires application of a demosaicing
algorithm that uses the available image data to estimate the values of
the two cone/sensor classes not present at each cone/sensor location. 
In this talk, I will review psychophysics and modeling that sheds light
on the demosaicing algorithm employed by the human visual system.  This
algorithm requires that the visual system have knowledge of the
spectral type of each of its cones.  For the L and M cones, a variety
of lines of evidence suggest that the class of the cone at each retinal
location is learned, rather than signaled by some sort of biochemical
marker.  In the second part of the talk, I will present results that
show that natural images contain sufficient statistical structure to
support unsupervised learning of cone classes.


- Learn More

David Brainard received his AB in physics from Harvard University
(1982) and MS (electrical engineering) and PhD (psychology) from
Stanford University in 1989. He is currently Professor of Psychology
at the University of Pennsylvania and his research focuses on human
color vision and color image processing. He is a fellow of the Optical
Society of America and the Association for Psychological Science.


October 7, 2011
11:00 am - 12:00 pm
Event Category: