TY - GEN
T1 - Accurate visual tracking by combining Bayesian and evolutionary optimization framework
AU - Jati, Grafika
AU - Gunawan, Alexander A.S.
AU - Jatmiko, Wisnu
AU - Febrian, Andreas
N1 - Publisher Copyright:
© 2016 IEEE.
PY - 2017/3/6
Y1 - 2017/3/6
N2 - Visual tracking is the process of locating, identifying, and determining of an object within video frames. From a Bayesian perspective, this is done by estimating the posterior density function. On the other hand, evolutionary optimization perspective would like to generate and select sufficiently optimize solution using two major components: diversification and intensification. This research will develop visual tracking algorithm using a Bayesian approach with evolutionary optimization in order to perform accurate tracking. The main idea is to combine Particle Markov Chain Monte Carlo (Particle-MCMC) as representation of Bayesian approach, with evolutionary optimization that is Particle Swarm Optimization (PSO) in each video frame. The visual tracking is regulated by Particle-MCMC filter algorithm and PSO will work within this filter to get more accurate tracking. Based on the dataset groundtruth, we found the accuracy of tracking can be increased considerably comparing to our previous research.
AB - Visual tracking is the process of locating, identifying, and determining of an object within video frames. From a Bayesian perspective, this is done by estimating the posterior density function. On the other hand, evolutionary optimization perspective would like to generate and select sufficiently optimize solution using two major components: diversification and intensification. This research will develop visual tracking algorithm using a Bayesian approach with evolutionary optimization in order to perform accurate tracking. The main idea is to combine Particle Markov Chain Monte Carlo (Particle-MCMC) as representation of Bayesian approach, with evolutionary optimization that is Particle Swarm Optimization (PSO) in each video frame. The visual tracking is regulated by Particle-MCMC filter algorithm and PSO will work within this filter to get more accurate tracking. Based on the dataset groundtruth, we found the accuracy of tracking can be increased considerably comparing to our previous research.
UR - http://www.scopus.com/inward/record.url?scp=85016931762&partnerID=8YFLogxK
U2 - 10.1109/ICACSIS.2016.7872795
DO - 10.1109/ICACSIS.2016.7872795
M3 - Conference contribution
AN - SCOPUS:85016931762
T3 - 2016 International Conference on Advanced Computer Science and Information Systems, ICACSIS 2016
SP - 523
EP - 528
BT - 2016 International Conference on Advanced Computer Science and Information Systems, ICACSIS 2016
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 8th International Conference on Advanced Computer Science and Information Systems, ICACSIS 2016
Y2 - 15 October 2016 through 16 October 2016
ER -