Advertisement

Camcorder Operation Judgement Using a Neural Computing Approach

  • Kitahiro Kaneda
  • Paul Wang
Part of the International Series in Intelligent Technologies book series (ISIT, volume 3)

Abstract

In this study, the neural computing approach is applied to the judgment of the camcorder operations which include fix-shot, panning and tilting. The data used in this study is the moving vector obtained from the real camcorder image. Also, seven features are carefully selected as possible inputs to this neural network. As a result of this approach, 96.88% accuracy of the operation judgment is obtained by using the length, angle of the moving vector and the time variation of the moving vector length as a set of three features. This result indicates that the approach proposed here can indeed classify three classes in the camcorder operations and it is also very easy to be utilized to control the automatic functions in the camcorder as well.

Keywords

Field Number Automatic Function Hide Layer Output Layer Layer Hide Layer Output Layer Satisfying Ration 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Sekine, T. Kondou, and H. Hrose, “Motion vector detecting system for video image stabilizers,” Proceedings of IEEE 1994 International Conference on Consumer Electronics (Digest of Technical Papers), pp.268–269, 1994.Google Scholar
  2. [3]
    Minsky and S. Papert, Perceptrons: An Introduction to Computational Geometry, MIT Press, Cambridge, Mass., 1969.MATHGoogle Scholar
  3. [4]
    J. Hopfield, “Neural networks and physical systems with emergent collective computational abilities,” Proceedings of Natl. Sci. USA, Vol. 323, pp. 533–536, 1986.Google Scholar
  4. [5]
    E. Rumelhart, G.E. Hinton, and R.J. Williams, “Learning representations by back-propagating errors,” Nature, Vol. 323, pp. 533–536, 1986.CrossRefGoogle Scholar

Copyright information

© Kluwer Academic Publishers 1995

Authors and Affiliations

  • Kitahiro Kaneda
    • 1
  • Paul Wang
    • 1
    • 2
  1. 1.Video Products Development Center and Fuzzy Logic Research LaboratoryCANON INC.TokyoJapan
  2. 2.Department of Electrical EngineeringDuke UniversityDurhamUSA

Personalised recommendations