Handwritten Character Recognition Based on Improved Convolutional Neural Network

Xue, Yu; Tong, Yiling; Yuan, Ziming; Su, Shoubao; Słowik, Adam; Toglaw, Sam

doi:10.32604/iasc.2021.016884

Cited by 6 publications

(3 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Accuracy Character Type Zhao, H., Liu, H. [10] 98% Digits Bora [17] 97.71% Digits Albahli [18] 99.78% Digits Hamida [19] 99.74% Digits Mhalgi [20] 86.0% Digits+Alphabets Yu Xue, Yiling Tong [21]…”

Section: Methodsmentioning

confidence: 99%

Design and Development of robust OCR system for hand filled form in Office Automation

Agrawal,

Roy,

Rakesh

et al. 2024

Preprint

View full text Add to dashboard Cite

Hand-filled forms are a ubiquitous means of information collection, yet their manual processing remains labor-intensive and error-prone. In this paper we proposed a novel approach to completely automate the processing of template based hand filled form and store the recognized data in central database. The proposed framework combines image processing techniques for form segmentation and a convolution neural network model, for accurate character recognition. By seamlessly integrating these components, the proposed approach streamlines the transition from handwritten content to digital data, reducing manual effort and potential errors.

show abstract

Section: Methodsmentioning

confidence: 99%

Design and Development of robust OCR system for hand filled form in Office Automation

Agrawal,

Roy,

Rakesh

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Tab. 4 shows the text recognition accuracy results of different methods, where '-' indicates that the recognition accuracy of the corresponding method is missing. For CASIA-HWDB dataset, the accurate rates of ResNet-26 and OrigamiNet-12 are 79.25% and 81.72% respectively, and our method achieve the most accurate rate is 90.53%, and acquire an improvement for the first two with 14.23% and 10.78% respectively.…”

Section: Recognition Accuracymentioning

confidence: 99%

“…From the initial single character recognition [4,5] to the current mainstream text line recognition [6][7][8], the field of HCTR has observed tremendous progresses for the past several decades, and text recognition has becoming a development trend reducing explicit segmentation proposal in conducive to increase multicharacters sequence recognition. There are many approaches learn to both simultaneously segment and recognize a handwritten text image representing a sequence of observations [3,[9][10][11].…”

Section: Introductionmentioning

confidence: 99%

End-to-end Handwritten Chinese Paragraph Text Recognition Using Residual Attention Networks

Wang¹,

Yang²,

Chen³

et al. 2022

Intelligent Automation &Amp; Soft Computing

View full text Add to dashboard Cite

Handwritten Chinese recognition which involves variant writing style, thousands of character categories and monotonous data mark process is a longterm focus in the field of pattern recognition research. The existing methods are facing huge challenges including the complex structure of character/line-touching, the discriminate ability of similar characters and the labeling of training datasets. To deal with these challenges, an end-to-end residual attention handwritten Chinese paragraph text recognition method is proposed, which uses fully convolutional neural networks as the main structure of feature extraction and employs connectionist temporal classification as a loss function. The novel residual attention gate block is more helpful in extracting essential features and making the training of deep convolutional neural networks more effective. In addition, we introduce the operations of batch bilinear interpolation which implement the mapping of two dimension text representation to one dimension text line representation without any position information of characters or text lines, and greatly reduce the labeling workload in preparing training datasets. In experimental, the proposed method is verified with two widely adopted handwritten Chinese text datasets, and achieves competitive results to the current state-of-the-art methods. Without using any position information of characters and text line, an accuracy rate of 90.53% is obtained in CASIA-HWDB test set.

show abstract

A Handwritten Number Recognition Scheme Based on Improved Convolutional Neural Network Algorithm

Li,

Pan,

Chen

et al. 2022

Communications in Computer and Information Science

View full text Add to dashboard Cite

Handwritten Character Recognition Based on Improved Convolutional Neural Network

Cited by 6 publications

References 15 publications

Design and Development of robust OCR system for hand filled form in Office Automation

Design and Development of robust OCR system for hand filled form in Office Automation

End-to-end Handwritten Chinese Paragraph Text Recognition Using Residual Attention Networks

A Handwritten Number Recognition Scheme Based on Improved Convolutional Neural Network Algorithm

Contact Info

Product

Resources

About