Skip navigation


  • DSpace logo
  • Home
  • Collections
  • Researcher Profile
  • Explore by
    • Researcher Profile
  • VGU library
  • Help
  • User Guide
  • Sign on to:
    • My DSpace
    • Receive email
      updates
    • Edit Account details

VGU RESEARCH REPOSITORY


Please use this identifier to cite or link to this item: https://epub.vgu.edu.vn/handle/dlibvgu/1632
DC FieldValueLanguage
dc.contributor.authorTran Hoang Nhat Khanhen_US
dc.date.accessioned2023-09-05T14:22:33Z-
dc.date.available2023-09-05T14:22:33Z-
dc.date.issued2023-
dc.identifier.urihttps://epub.vgu.edu.vn/handle/dlibvgu/1632-
dc.description.abstractOptical Character Recognition (OCR) is a topic that has been studied over the past few decades with various applications. The project aims to develop and implement automated OCR software using a multi-layer neural network with a Long Short-Term Memory (LSTM) method. By using this method, it can recognize optical characters from the given input image with high accuracy. OCR problem could be simplified as a classification problem, and like any typical classification problem, it has training and testing steps. With neural networks, OCR systems will be learned with a supervised learning algorithm by using the training examples to teach the networks how to classify them into different classes. The number of classes depends on how many characters need to be classified. When training, all the examples will be displayed on a feature space and can be grouped by a hyperplane. This hyperplane is a boundary line to separate the classes into the decision planes. The accuracy of the decision plane depends on the value of the weight and the bias of the corresponding neurons. The classification now is turning to the minimization problem where the optimal of the weights and the bias need to be found to ensure the accuracy of the decision planes. Stochastic gradient descent method is a prospective method that could minimize the cost function with a limited training example called mini-batch. The advantage of the stochastic gradient descent method is memoryless and time-saving because instead of using all 60,000 training examples, this method will choose a mini-batch with only 100 examples or less.en_US
dc.language.isoenen_US
dc.rightsAttribution-NonCommercial 4.0 International*
dc.rights.urihttps://creativecommons.org/licenses/by-nc/4.0/*
dc.subjectNeural networksen_US
dc.subjectOptical character recognitionen_US
dc.titleOptical character recogition using neutral networksen_US
dc.typeThesisen_US
item.fulltextWith Fulltext-
item.languageiso639-1other-
item.grantfulltextrestricted-
Appears in Collections:Electrical Engineering and Information Technology (EEIT)
Files in This Item:
File Description SizeFormat Existing users please Login
Optical character recogition using neutral networks.pdf11.38 MBAdobe PDF
Show simple item record

Page view(s)

104
checked on Aug 23, 2025

Download(s)

20
checked on Aug 23, 2025

Google ScholarTM

Check


This item is licensed under a Creative Commons License Creative Commons

© Copyright 2020 by Vietnamese - German University Library.
Add: Ring road 4, Quarter 4, Thoi Hoa Ward, Ben Cat City, Binh Duong Province
Tel.:(0274) 222 0990. Ext.: 70206