Department of Information and Computing Sciences

Departement Informatica Onderwijs
Bachelor Informatica Informatiekunde Kunstmatige intelligentie Master Computing Science Game&Media Technology Artifical Intelligence Business Informatics

Onderwijs Informatica en Informatiekunde

Vak-informatie Informatica en Informatiekunde

Geometric methods in shape and pattern recognition

Website:website containing additional information
Course code:INFOGMSPR
Credits:7.5 ECTS (=5.25 old credit points)
Period:period 3 (week 6 through 16, i.e., 6-2-2006 through 21-4-2006; retake week 21)
Timeslot:C
Participants:up till now 8 subscriptions
Schedule:Dit is een oud rooster!
formgrouptimeweekroomteacher
coordinatie          Remco Veltkamp
 
seminar   Mon 13-156-11,13-15 BBL-475 prof.dr.H.Alt
  
Tue 15-179-11, 13-15 BBL-475
Thu 11-136-11,13-15 BBL-475
Contents:In recent years there has been considerable research work on applying methods from computational geometry to measuring and detecting similarities in patterns and shapes, which are modelled by sets of points, line segments or other curves. The seminar will consider the most important results in that field, covering the following topics.
  • Basics from computational geometry (mainly Voronoi diagrams of points, line segments, and curves).
  • Patterns, shapes, transformations, and distance measures.
  • Matching by Hausdorff distance.
  • Reference points.
  • Measuring the Frechet distance (includes parametric search).
  • Dynamic time warping.
  • Probabilistic matching.
  • Higher dimensional matching.
  • Pattern and shape retrieval: vantage point methods, nearest neighbor in high-dimensional spaces.
  • Analytic approaches to describe and match shapes.
  • Applications.
  • Simplification of shapes, detecting symmetries.
Literature:Handouts.
Course form:Presentations by the instructor and the participants, small weekly homeworks, further details to be announced.
Minimum effort to qualify for 2nd chance exam:Will be announced the first meeting.
Description:In recent years there has been considerable research work on applying methods from computational geometry to measuring and detecting similarities in patterns and shapes, which are modelled by sets of points, line segments or other curves. The seminar will consider the most important results in that field, covering the topics described above.

Prerequisite: course in algorithms and data structures.

wijzigen?