2013
DOI: 10.1016/j.aasri.2013.10.045
|View full text |Cite
|
Sign up to set email alerts
|

Off-line Handwritten Character Recognition Using Features Extracted from Binarization Technique

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
14
0

Year Published

2015
2015
2023
2023

Publication Types

Select...
6
4

Relationship

0
10

Authors

Journals

citations
Cited by 51 publications
(14 citation statements)
references
References 4 publications
0
14
0
Order By: Relevance
“…In 2007 Manjunath Aradhya et al [13] showed a multi-lingual OCR leading to a good accuracy. Afterwards, Desai [14] used OCR technique for recognizing Gujrati handwritten digits in 2010 resulting in 82 % success rate and likewise Choudhary et al [15] routine this with binarization method to judge capability of OCR for english characters in 2013 giving accuracy of 85.62 %.…”
Section: Related Workmentioning
confidence: 98%
“…In 2007 Manjunath Aradhya et al [13] showed a multi-lingual OCR leading to a good accuracy. Afterwards, Desai [14] used OCR technique for recognizing Gujrati handwritten digits in 2010 resulting in 82 % success rate and likewise Choudhary et al [15] routine this with binarization method to judge capability of OCR for english characters in 2013 giving accuracy of 85.62 %.…”
Section: Related Workmentioning
confidence: 98%
“…This technique yields ~90% accuracy for different character seeking angles but it was computationally expensive as feature extraction proves to be time consuming. www.ijacsa.thesai.org Amit Choudhary et al [7] presented an extensive work on offline handwritten English character recognition using multilayer feed forward neural network reporting an accuracy of 85.62%. Ganai et al [8] focused on improving the speech quality of existing system by combining the methods of Hidden Markov Model (HMM)-based speech synthesis and waveform-based speech synthesis to develop human like speech.…”
Section: IImentioning
confidence: 99%
“…Such forward network with hidden layer(s) has an important feature that the units in the hidden layer(s) can randomly form their own input presentation and the weights between the units of the input layer and the hidden layer determines whether each unit in the hidden layer is activated. BP algorithm is slow in learning and the predictive ability training capability (learning capability) of its network is contradictory (Parshuram and Ravinda, 2015;Amit and Rahul et al, 2013). The structure of BP network is shown as Fig.7.…”
Section: Bp Algorithmmentioning
confidence: 99%