BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Image Visualization and Restoration by Curvature Motions - Ciomaga
 \, A (cole Normale Suprieure)
DTSTART:20110822T150000Z
DTEND:20110822T153000Z
UID:TALK32430@talks.cam.ac.uk
CONTACT:Mustapha Amrani
DESCRIPTION:The role of curvatures in visual perception goes back to '54 a
 nd is due to Attneave. It can be argued on neurological grounds that human
  brain could not possible use all the information provided by states of si
 mulation. But information that stimulates the retina\, is located at regio
 ns where color changes abruptly (contours)\, and furthermore at angles and
  peaks of curvature. Yet\, a direct computation of curvatures on a raw ima
 ge is impossible. \nWe show in this presentation how curvatures can be acc
 urately estimated\, at subpixel resolution\, by a direct computation on le
 vel lines after their independent smoothing. This view towards shape analy
 sis requires a representation of an image in terms of its level lines. At 
 the same time\, it involves short time smoothing (in occurrence Curve Shor
 tening or Af?ne Shortening) simultaneously for level lines and images. \nI
 n this setting\, we found an explicit connection between the geometric app
 roach for Curve / Af?ne Shortening and the viscosity approach for the Mean
  / Af?ne Curvature Motion\, based on a complete image processing pipeline\
 , that we term Level Lines (Af?ne) Shortening\, shortly LL(A)S. We show th
 at LL(A)S provides an accurate visualization tool of image curvatures\, th
 at we call an Image Curvature Microscope. As an application we give some i
 llustrative examples of image visualization and restoration: noise\, JPEG 
 artifacts\, and aliasing will be shown to be nicely smoothed out by the su
 bpixel curvature motion.\n\n
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR