Restoring Images Taken from Scratched 2-Inch Tape

  • Steven Armstrong
  • Peter J. W. Rayner
  • Anil C. Kokaram


This paper presents a method of restoring video images taken from scratched 2-inch video tape. Because of the way that the video tracks are laid down, a scratch along this tape causes severe degradation. The paper adopts a detect and repair strategy, the detection being accomplished automatically with a rank order model-based approach that makes use of the EM algorithm. The detection is improved by exploiting the known periodicity of the degradation artefacts. Results demonstrate the method’s improved preservation of image features over previous methods, while retaining a high percentage of correct detections.


Correct Detection Laplacian Distribution Median Prediction Video Tape Recorder Important Image Feature 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Christopher M. Bishop. Neural Networks for Pattern Recognition. Oxford University Press, 1995.Google Scholar
  2. [2]
    N.R. Harvey and S. Marshall. Application of non-linear image processing: Digital video restoration. In IEEE International Conference on Image Processing (ICIP), volume I, pages 731–734, 1997.Google Scholar
  3. [3]
    David Kirk. Audio and Video Recording. Faber & Faber, 1975.Google Scholar
  4. [4]
    Charles W. Therrian. Discrete Random Signals and Statistical Signal Processing. Prentice-Hall International Editions, 1992.Google Scholar
  5. [5]
    Gordon White. Video Recording: Record and Replay Systems. Newnes-Butterworths, 1972.Google Scholar

Copyright information

© Springer-Verlag London Limited 1998

Authors and Affiliations

  • Steven Armstrong
    • 1
  • Peter J. W. Rayner
    • 1
  • Anil C. Kokaram
    • 2
  1. 1.Signal Processing and Communications LaboratoryCambridge University Engineering DepartmentCambridgeEngland
  2. 2.Trinity CollegeDublinIreland

Personalised recommendations