1. Science
  2. Publications
  3. Information Processing Systems
  4. 2(92)'2011
  5. Criteria and models of structural classification using voting principle

Criteria and models of structural classification using voting principle

V.O. Gorohovatsky, T.V. Polyakova
Annotations languages:

The ways of constructing models and criteria for classification of video objects structural descriptions based on voting are discussed. The construction system of classification with high immunity to additive and spatial interference is offered by selection models for structural similarity measures, given their parametric definition. Results of experiments are analyzed.
Keywords: image classification, structurally-hierarchical methods, characteristic features, structural description, voting principle, criterion of classification, model of comparison, reliability of classification