A Fuzzy Reasoning Technique for Pattern Recognition
| Autor: | Manjaiah D. H, Ashoka H N, Rabindranath Bera |
|---|---|
| Rok vydání: | 2012 |
| Předmět: |
Fuzzy classification
Computer science business.industry Binary image Feature extraction Pattern recognition Optical character recognition computer.software_genre Numeral system ComputingMethodologies_PATTERNRECOGNITION Knowledge base Computer Science::Computer Vision and Pattern Recognition Fuzzy reasoning Data mining Artificial intelligence business Classifier (UML) computer |
| Zdroj: | International Journal of Computer Applications. 49:19-23 |
| ISSN: | 0975-8887 |
| DOI: | 10.5120/7679-0980 |
| Popis: | This paper presents a fuzzy reasoning technique for pattern classification. The binary image of a pattern is partitioned into number of overlapping and non-overlapping regions for feature extraction. A set of features extracted to represent the binary images are used in the classifier design. The knowledge base (KB) is build using the statistical information involved in the interclass region features. The unknown samples features are matched with KB and assigned a membership value by the fuzzy reasoning technique for their classification. On experimentation with the handwritten numeral data, found good classification and recognition rate. General Terms Pattern Recognition, Fuzzy classification, Optical Character Recognition |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|