Abstract
A neural network approach to the extraction of 1hop F-layer traces from oblique-incidence ionograms is shown to offer performance at least comparable with conventional filtering techniques. Preprocessing in the form of background noise and vertical (horizontal) line removal is utilised prior to training a 110∶7∶100 MLP using backpropagation with momentum. It is further demonstrated that such successful trace extraction can be achieved with just 50 ionogram training exemplars.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Coleman C. Jornoix Software. DSTO, Adelaide, 1996.
Simon W, Carter J. Generalizing from a small set of training exemplars for handwritten digit recognition. In SK Rogers (ed) Applications of Artificial Neural Networks II. Proc. SPIE 1991; 1469: 592–601.
Yan H. Building a robust nearest neighbour classifier containing only a small number of prototypes. Int J Neural Systems 1992; 3(4): 361–369.
Fulcher J. Archaeological provenancing. In E Fiesler and R Beale (eds) Handbook of Neural Computation. Institute of Physics/Oxford University Press, 1997.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Hagenbuchner, M., Fulcher, J. Noise removal in ionograms by neural network. Neural Comput & Applic 6, 165–172 (1997). https://doi.org/10.1007/BF01413828
Issue Date:
DOI: https://doi.org/10.1007/BF01413828