Optical Character Recognition using Neural Network
Abstract
Character Recognition (CR) has always been an active research field in the past and is still a challenging research topic due to its wide range of applications. The most important methods of offline handwriting recognition can be divided into two categories: overall and segmentation- based.
When considering the global features extracted from the complete word image, the holistic method is used when recognizing limited vocabulary. On the other hand, the segment-based strategy uses a bottom-up approach, starting at the line or symbol level, and processing meaningful words. After segmentation, the problem comes down to recognizing a single character or line. Therefore, the system can use unlimited vocabulary.
Full Text:
PDFReferences
Faustino Gomez, Jrgen Schmidhuber, Santiago Fernndez, Alex Graves, Connectionist of the temporal classification: labelling of unsegmented sequence of data with the recurrent of neural networks,the Proceedings of the 23rd international conference on the Machine learning. 2006
The Convolutional Neural Network written Benchmarks: cnn-benchmarks
Elliot Cuzzillo, Elie Krevat. Improving the Off-line Handwritten Character Recognition with Hidden Markov Models.
Fabian Tschopp. Efficiently given Convolutional Neural Networks for the Pixelwise Classification on the Heterogeneous ‘ Hardware Systems
George Nagy’ The Document processing applications.
Mail encoding and processing system patent: https://www.google.com/patents/US5420403
Kurzweil Computer Products. http://www.kurzweiltech.com/kcp.html
H. Bunke1, M. Roth1, E.G. Schukat- Talamazzini. Offline Cursive Handwriting Recognition using Hidden Markov Models.
K. Simonyan, A. Zisserman Very Deep Convolutional Networks for Large-Scale Image Recognition arXiv technical report, 2014
Lisa Yan. Recognizing Handwritten Characters. CS 231N Final Research Paper.
Refbacks
- There are currently no refbacks.
This work is licensed under a Creative Commons Attribution 3.0 License.