Scholarlink

Deep learning methodologies are used for the recognition of handwritten Devanagari characters. The Devanagari script is widely used across Indian languages comprises 47 letters, 14 vowels, 34 consonants and also consists of 10 digits. Deep learning model with Convolutional Neural Network is developed and fine-tuned to effectively recognize and classify Devanagari characters, and this model is trained on a handwritten Devanagari characters dataset consisting of 92,000 images. The experimental results show the model’s accuracy and efficiency in identifying characters of various writing styles. The output and results processed by this deep Learning model will contribute in the advancements of Devanagari character recognition and also able to extend its application for solving complex problems. The potential applications of this model can extend for document analysis, language processing, and cultural heritage preservation.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparative Analysis and Implementation of CNN-Based Deep Learning Model (DHCRnet) for Devanagari Handwritten Character Recognition

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Sanket Ujgare,
  • Vaibhav Chaudhari,
  • Saurabh Dhavale,
  • Rohit Dadhel,
  • Ram Joshi,
  • Dhanjraj Jadhav

摘要

Deep learning methodologies are used for the recognition of handwritten Devanagari characters. The Devanagari script is widely used across Indian languages comprises 47 letters, 14 vowels, 34 consonants and also consists of 10 digits. Deep learning model with Convolutional Neural Network is developed and fine-tuned to effectively recognize and classify Devanagari characters, and this model is trained on a handwritten Devanagari characters dataset consisting of 92,000 images. The experimental results show the model’s accuracy and efficiency in identifying characters of various writing styles. The output and results processed by this deep Learning model will contribute in the advancements of Devanagari character recognition and also able to extend its application for solving complex problems. The potential applications of this model can extend for document analysis, language processing, and cultural heritage preservation.