A Survey of Methods and Strategies in Character Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
A New Methodology for Gray-Scale Character Segmentation and Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
On-Line Hand-Printing Recognition with Neural Networks
MICRONEURO '96 Proceedings of the 5th International Conference on Microelectronics for Neural Networks and Fuzzy Systems
ICPR '02 Proceedings of the 16 th International Conference on Pattern Recognition (ICPR'02) Volume 3 - Volume 3
IWFHR '02 Proceedings of the Eighth International Workshop on Frontiers in Handwriting Recognition (IWFHR'02)
Handwritten Numeral String Recognition Using Neural Network Classifier Trained with Negative Data
IWFHR '02 Proceedings of the Eighth International Workshop on Frontiers in Handwriting Recognition (IWFHR'02)
ICASSP '01 Proceedings of the Acoustics, Speech, and Signal Processing, 2001. on IEEE International Conference - Volume 03
Handwritten word recognition with character and inter-character neural networks
IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
An Approach to Extracting the Target Text Line from a Document Image Captured by a Pen Scanner
ICDAR '03 Proceedings of the Seventh International Conference on Document Analysis and Recognition - Volume 1
| Hi-index | 0.00 |
In the past several years, we've been developing ahigh performance OCR engine for machine printed Chinese/English documents. We have reported previously (1)how to use character modeling techniques based on MCE(minimum classification error) training to achieve the highrecognition accuracy, and (2) how to use confidence-guidedprogressive search and fast match techniques to achieve thehigh recognition efficiency. In this paper, we present twomore techniques that help reduce search errors and improvethe robustness of our character recognizer. They are (1)to use MCE-trained character-pair models to avoid error-pronecharacter-level segmentation for some trouble cases,and (2) to perform a MCE-based negative training to improvethe rejection capability of the recognition models onthe hypothesized garbage images during recognition process.The efficacy of the proposed techniques is confirmedby experiments in a benchmark test.