Improving Kannada OCR Using a Stroke-Based Approach
| Autor: | Koshy George, Chakith Pattar, Edupuganti Arun, J. Vinith |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Sequence
Computer science business.industry 05 social sciences Feature extraction ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Pattern recognition 010501 environmental sciences 01 natural sciences Image (mathematics) Task (project management) Support vector machine Character (mathematics) 0502 economics and business Task analysis Artificial intelligence 050207 economics business 0105 earth and related environmental sciences |
| Zdroj: | TENCON |
| DOI: | 10.1109/tencon.2019.8929606 |
| Popis: | This paper presents a method of utilizing generated strokes as additional information to perform the task of Character Recognition. We propose a model which generates a sequence of strokes given the character, which when provided along with the complete image, improves the recognition. From our experiments, we observe that this approach makes the model better equipped for a noisy environment, and tolerant to cracks or gaps in the character. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|