Campus Event Calendar
Event Entry
New for: D2, D3
What and Who
Light in a Box: Ye Olde Camera in a Different Perspective
Ivo Ihrke
Max-Planck-Institut für Informatik - D4
Joint MPI-INF/MPI-SWS Lecture Series
AG 1, AG 2, AG 3, AG 4, AG 5, SWS, RG1, MMCI
MPI Audience
English
Note: We use this to send email in the morning.
Date, Time and Location
Wednesday, 1 December 2010
16:00
60 Minutes
E1 4
024
Saarbrücken
Abstract
My talk will cover recent advances in computational photography as
well as the underlying conceptual change in camera models for computer
graphics and computer vision. I will argue that the current
development has, as of now, not considered the general properties of
generalized computational imaging systems such as sampling rates,
reconstruction potential and properties of the underlying function
that is to be captured.
I present some recent ideas on a generalized imaging model for future
cameras that could serve as a tool to answer these questions. Instead
of just considering a single ray per pixel, I propose to look at the
manifold of directional, spectral and temporal variation that is being
integrated by a single sensel in our current imaging sensors. While
the arrangement of these sensels and the optical systems used could be
changed in future designs, the concept of a sensel as an integrating
device will most likely persist. This view yields a conceptually
simple model of computational imaging devices as projections onto basis
functions that every sensel defines. These bases can represent certain
subspaces of the plenoptic function.
I present some recent results of analyzing this model for the more
restricted setup of sensels laid out on a two-dimensional cartesian
grid. I conclude by discussing the promises and challenges of the
proposed model.
well as the underlying conceptual change in camera models for computer
graphics and computer vision. I will argue that the current
development has, as of now, not considered the general properties of
generalized computational imaging systems such as sampling rates,
reconstruction potential and properties of the underlying function
that is to be captured.
I present some recent ideas on a generalized imaging model for future
cameras that could serve as a tool to answer these questions. Instead
of just considering a single ray per pixel, I propose to look at the
manifold of directional, spectral and temporal variation that is being
integrated by a single sensel in our current imaging sensors. While
the arrangement of these sensels and the optical systems used could be
changed in future designs, the concept of a sensel as an integrating
device will most likely persist. This view yields a conceptually
simple model of computational imaging devices as projections onto basis
functions that every sensel defines. These bases can represent certain
subspaces of the plenoptic function.
I present some recent results of analyzing this model for the more
restricted setup of sensels laid out on a two-dimensional cartesian
grid. I conclude by discussing the promises and challenges of the
proposed model.
Contact
Jennifer Müller
900
--email hidden
passcode not visible
logged in users only
Jennifer Müller, 11/29/2010 12:36
Jennifer Müller, 08/26/2010 09:35
Anna-Lisa Overhoff, 08/26/2010 09:23 -- Created document.
Jennifer Müller, 08/26/2010 09:35
Anna-Lisa Overhoff, 08/26/2010 09:23 -- Created document.