Nearness of subtly different digital images

  • Authors:

  • Leszek Puzio;James F. Peters

  • Affiliations:

  • Computational Intelligence Laboratory, Department of Electrical & Computer Engineering, University of Manitoba, Winnipeg, Manitoba, Canada,Department of Information Systems and Applications, U ...;Computational Intelligence Laboratory, Department of Electrical & Computer Engineering, University of Manitoba, Winnipeg, Manitoba, Canada,School of Mathematics & Computer/Information Scie ...

  • Venue:
  • Transactions on Rough Sets XVI
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

The problem considered in this article is how to measure the nearness or apartness of digital images in cases where it is important to detect subtle changes in the contour, position, and spatial orientation of bounded regions. The solution of this problem results from an application of anisotropic (direction dependent) wavelets and a tolerance near set approach to detecting similarities in pairs of images. A wavelet-based tolerance Nearness Measure (tNM) makes it possible to measure fine-grained differences in shapes in pairs of images. The application of the proposed method focuses on image sequences extracted from hand-finger movement videos. Each image sequence consists of hand-finger movements recorded during rehabilitation exercises. The nearness of pairs of images from such sequences is measured to check the extent that normal hand-finger movement differs from arthritic hand-finger movement. Experimental results of the proposed approach are reported, here. The contribution of this article is an application of an anisotropic wavelet-based tNM in classifying arthritic hand-finger movement images in terms of their degree of nearness to or apartness from normal hand-finger movement images.