CNCSIS IDEI, Fuzing Statistics and Semantic Modeling in Image Sequences Analysis
ID_930, 667/19.01.2009
Probabilistic approaches to image sequence analysis have difficulty modeling the complex situation encountered in real-world applications. To alleviate this problem, we propose a new theoretical framework for merging the level of statistical thinking with the level of semantics for the benefit of both levels. We will test the effectiveness of the concept of object tracking and motion task estimation related to human body motion analysis.
FINANCED PROJECT
We define three objection main research:
- The first of them is the development of a semantically guided Traker core. The best method to exploit semantic information extracted from the image sequence by improving Traking performance will be investigated.
- The second research goal is to find an efficient use of the new sparse representation in motion modeling and semantic inference.
- The third research objective is to strengthen a fund of higher level information extracted from the processed image sequence.
The main result:
- Development of a semantically guided tracker;
- Estimation of the robust fund
PROJECT DATA
Director
Prof. dr. ing Vasile GUI
Value
167,831 RON
States
Prof. Dr. Eng. Florin ALEXA, Assoc.prof.dr.eng. Cătălin CALEANU, Teach.assist.dr.eng. Ciprian DAVID, Teach.assist.eng. Gheorghe POPA, Dr. Eng. Georgiana SIMION