Handwriting recognition using high performance computing platforms
Investigation on the feasibility of various character features extracted for handwritten character recognition are comprehensively benchmarked and many variants classifiers using neural network technologies are described. Extensive coverage of preprocessing techniques are listed and illustrated. Neu...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2009
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/20433 |
| Summary: | Investigation on the feasibility of various character features extracted for handwritten character recognition are comprehensively benchmarked and many variants classifiers using neural network technologies are described. Extensive coverage of preprocessing techniques are listed and illustrated. Neural network technologies in particular, the backpropagation neural network are extensively reviewed and studied in the context of learning problems in handwriting recognition. The use of post-processor techniques and multi-neural network architectures in character recognition are also presented. |
|---|