Elsevier

Pattern Recognition Letters

Volume 15, Issue 2, February 1994, Pages 105-109
Pattern Recognition Letters

Winograd's method: a perspective for some pattern recognition problems

https://doi.org/10.1016/0167-8655(94)90039-6Get rights and content

Abstract

In this article, Winograd's method is used with Euclidean distance, Mahalanobis distance and Maximum Likelihood classifiers to reduce their computational time requirements. Experimental work is carried out with 6 band thematic mapper data. The proposed fast algorithms for Euclidean and Maximum Likelihood classifiers are observed to be 2 times faster than their literal algorithms. In the case of the Mahalanobis distance classifier, the proposed fast algorithm is showing a speed-up of 7. Use of this logic with other pattern recognition algorithms is also discussed.

References (7)

There are more references available in the full text version of this article.

Cited by (4)

  • A fast two-stage classification method for high-dimensional remote sensing data

    1998, IEEE Transactions on Geoscience and Remote Sensing
  • Winograd's Inequality: Effectiveness for efficient training of Deep Neural Networks

    2018, International Journal of Intelligent Systems and Applications
  • Various approaches to speed up mahalanobis distance classifier

    1995, International Journal of Remote Sensing
View full text