Palestra: Rede de Sensores sem fio para Monitorar o meio Ambiente.
Document Actions
Palestra: Novel Feature Qualities for Tracking.
Prof. Dr. Michael Eckmann - Mathematics and Computer Science Department Skidmore College, Saratoga Springs, NY, USA, na Série de Seminários 2011 da Pós-Graduação, dia 03/06/2011, às 14:00 h, Auditório do IC, Sala 85 - IC 2.
| What | Palestra |
|---|---|
| When |
03/06/2011 from 14:00 to 15:00 |
| Where | Auditório do IC - Sala 85 - IC 2 |
| Add event to calendar |
vCal
iCal
|
Two new and important qualities of features which expand upon the usual local feature information are described. Spatio-Temporal consistency is a quality of a feature that quantifies how consistently a feature has been tracked over time and how smooth its motion was in space. Distributivity is a quality of a feature that quantifies physical distance (in number of pixels) from other features in the same frame. These qualities are applied to several widely used and well known approaches to feature detection — Shi and Tomasi’s Good Features to Track and Lowe’s Scale Invariant Feature Transform (SIFT). Results of comparisons of these feature detectors with and without the added qualities are described. Specifically, we show that these feature detectors enhanced with the qualities that we define improve matching. We also show how spatiotemporal consistency can handle occlusion and show evidence of how distributivity improves a mosaicing application. ================================================================= Michael Eckmann received the Ph.D. degree in computer science in 2007 from Lehigh University, Bethlehem, PA, USA. He also holds a Masters degree in computer science, 1999, a Bachelor of Science degree in computer engineering, 1990, and a Bachelor of Arts degree in mathematics, 1990, all from Lehigh University. Between undergraduate and graduate school he worked in industry for eight years. In the past he has taught computer related courses at Lehigh University and Wilkes University. Dr. Eckmann is currently an Assistant Professor in the Mathematics and Computer Science Department of Skidmore College, Saratoga Springs, NY, USA. He teaches courses in programming languages, computer graphics and computer vision, among others. His research interests are in computer vision and image processing, specifically feature tracking, surveillance, and most recently digital image forensics. He has authored/co-authored several publications in these areas. He is a member of ACM, IEEE and IEEE Computer Society. ================================================================= Organizadora: Profa. Beatriz Toledo (beatriz@ic.unicamp.br)
IC / Unicamp
Fone: (019) 3521-5869 =================================================================
VII WTD do IC 2012
