Skip to Main content Skip to Navigation
Conference papers

Pattern Classification with Rejection Using Cellular Automata-Based Filtering

Abstract : In this article we address the problem of contaminated data in pattern recognition tasks, where apart from native patterns we may have foreign ones that do not belong to any native class. We present a novel approach to image classification with foreign pattern rejection based on cellular automata. The method is based only on native patterns, so no knowledge about characteristics of foreign patterns is required at the stage of model construction. The proposed approach is evaluated in a study of handwritten digits recognition. As foreign patterns we use distorted digits. Experiments show that the proposed model classifies native patterns with a high success rate and rejects foreign patterns as well.
Complete list of metadata

Cited literature [14 references]  Display  Hide  Download

https://hal.inria.fr/hal-01656223
Contributor : Hal Ifip <>
Submitted on : Tuesday, December 5, 2017 - 2:57:32 PM
Last modification on : Wednesday, December 6, 2017 - 1:21:02 AM

File

448933_1_En_1_Chapter.pdf
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Agnieszka Jastrzebska, Rafael Toro Sluzhenko. Pattern Classification with Rejection Using Cellular Automata-Based Filtering. 16th IFIP International Conference on Computer Information Systems and Industrial Management (CISIM), Jun 2017, Bialystok, Poland. pp.3-14, ⟨10.1007/978-3-319-59105-6_1⟩. ⟨hal-01656223⟩

Share

Metrics

Record views

146

Files downloads

106