Plasma and Fusion Research

Volume 13, 3405117 (2018)

Regular Articles

Estimation of Plasma Emission Transition Using Hidden Markov Model
Shota NAKAGAWA1), Teruhisa HOCHIN1,2), Hiroki NOMIYA1) and Hideya NAKANISHI2)
1)
Kyoto Institute of Technology, Kyoto 606-8585, Japan
2)
National Institute for Fusion Science, Toki 509-5292, Japan
(Received 27 December 2017 / Accepted 8 August 2018 / Published 25 September 2018)

Abstract

This study proposes a method for estimating plasma-emission transitions from plasma-emission videos using a hidden Markov model (HMM). The proposed method retrieves similar videos and learns model parameters from them. The plasma-emission characteristics that we have employed are color, brightness, position, shape, and the speed at which the brightness of a plasma emissions changes. Multiple HMMs based on these plasma-emission characteristics are employed to represent the plasma-emission patterns. The anticipated plasma-emission transitions are estimated using state-transition probabilities from the generated model. Experimental results are used to confirm that the proposed methods are effective in identifying similar plasma videos and estimating probable future states of the plasma.

Keywords

hidden Markov model, video, plasma, similarity, estimation

DOI: 10.1585/pfr.13.3405117

Full Text

PDF format (3.0Mbytes) PDF Download

References

  • [1] Large Helical Device Project Home Page, http://www.lhd.nifs.ac.jp/ (reference 12-21-2017).
  • [2] M. Shoji, K. Yamazaki and S. Yamaguchi, “Development of the Real-time lmage Data Acquisition System for Observing the Plasma Dynamic Behavior of LHD Long-pulse Discharges”, J. Plasma Fusion Res. SERIES 3, 440 (2000).
  • [3] J. Bilmes, “A gentle tutorial on the EM algorithm and its application to parameter estimation for Gaussian mixture and hidden markov models” (Technical Report ICSI-TR-97-02, University of Berkeley, 1997).
  • [4] H. Woo, Y. Ji, H. Kono and Y. Tamura, “Estimation Method for Lane Changes of Other Traffic Participants Using State-Unit based Hidden Markov Models”, Robotics Symposia proceedings 21, 222 (2016).
  • [5] H. Ohya and S. Morishima, “Automatic Music Video Generating System by Learning Existing Contents Based on Hidden Markov Model”, Music and computer (MUS) Technical Report 2012-MUS-95, 1 (2012).
  • [6] H. Shiroshita, T. Hochin, H. Nomiya and H. Nakanishi, “Similarity Retrieval of Plasma Videos and Its Evaluation”, Proc. of 3rd Int'l Conf. on Applied Comp. and Inf. Tech. (ACIT 2015), 229 (2015).
  • [7] H. Shiroshita, T. Hochin, H. Nomiya and H. Nakanishi, “Revised Similarity Retrieval Method of Plasma Emission Videos and Its Evaluation”, Proc. of 18th IEEE/ACIS Int'l Conf. on Softw. Eng., Artif. Intell., Networking and Parallel/Distr. Comp. (SNPD 2017), 575 (2017).
  • [8] Y. Chida, H. Koga and T. Watanabe, “Fast Detection of Query Videos from Video Streams” (DEIM Forum 2013 D1-5, 2013).