Two-Step CNN Framework for Text Line Recognition in Camera-Captured Images
| Autor: | Vladimir V. Arlazarov, Alexander Sheshkus, Yulia S. Chernyshova |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
General Computer Science
Computer science character recognition Image processing 02 engineering and technology Overfitting 0502 economics and business 0202 electrical engineering electronic engineering information engineering General Materials Science Segmentation Text recognition character segmentation Artificial neural network business.industry 05 social sciences General Engineering Pattern recognition ComputingMethodologies_PATTERNRECOGNITION machine learning 020201 artificial intelligence & image processing Artificial intelligence lcsh:Electrical engineering. Electronics. Nuclear engineering Line (text file) business artificial neural networks lcsh:TK1-9971 050203 business & management |
| Zdroj: | IEEE Access, Vol 8, Pp 32587-32600 (2020) |
| ISSN: | 2169-3536 |
| Popis: | In this paper, we introduce an “on the device” text line recognition framework that is designed for mobile or embedded systems. We consider per-character segmentation as a language-independent problem and individual character recognition as a language-dependent one. Thus, the proposed solution is based on two separate artificial neural networks (ANN) and dynamic programming instead of employing image processing methods for the segmentation step or end-to-end ANN. To satisfy the tight constraints on memory size imposed by embedded systems and to avoid overfitting, we employ ANNs with a small number of trainable parameters. The primary purpose of our framework is the recognition of low-quality images of identity documents with complex backgrounds and a variety of languages and fonts. We demonstrate that our solution shows high recognition accuracy on natural datasets even being trained on purely synthetic data. We use MIDV-500 and Census 1961 Project datasets for text line recognition. The proposed method considerably surpasses the algorithmic method implemented in Tesseract 3.05, the LSTM method (Tesseract 4.00), and unpublished method used in the ABBYY FineReader 15 system. Also, our framework is faster than other compared solutions. We show the language-independence of our segmenter with the experiment with Cyrillic, Armenian, and Chinese text lines. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|