Devanagari

About: Devanagari is a research topic. Over the lifetime, 655 publications have been published within this topic receiving 7428 citations. The topic is also known as: Deva nagari & Hindi Script.

Proposal to Encode the Tirhuta Script in ISO/IEC 10646

Abstract: Author(s): Pandey, Anshuman | Abstract: This is a proposal to encode the Tirhuta script in the international character encoding standard Unicode. This script was published in Unicode Standard version 7.0 in June 2014. Tirhuta was the main writing system for the Maithili language, which is spoken in India and Nepal. The script continued in use until the late 20c, when Devanagari replaced it.

SVM-PCA based Handwritten Devanagari Digit Character Recognition

Abstract: The blended fusion of Support Vector Machine (SVM) and Principal Component Analysis (PCA) have been widely used in recognizing handwritten digit characters of Devanagari script. The feature information from the character is extracted using its skeleton structure which optimally reduce data dimensionality using PCA. There is ample information available on handwritten charac-ter recognition on Indian and Non-Indian scripts but very few article emphasized on recognition of Devanagari scripts. Therefore, this paper presents an efficient handwritten Devanagari character recognition system based on block based feature extraction and PCA-SVM classifier. We have collected samples of handwritten Devanagari characters from different handwritten experts for classification. For experimental work, total of 100 images having Devanagari digit characters been used for the purpose of training and testing. The proposed system achieves a maximum recognition accuracy of 96.6 % and 96.5% for 5 & 10 fold validations with 70% training and 30% testing data using block based feature and SVM classifier having different kernels. The obtained results achieve maximum accuracy using SVM classifier for digit character recognition. In future deep learning networks will be considered for accuracy enhancement and precision.

Touching character segmentation of Devanagari script

Abstract: Segmentation of characters is one of the major step in OCR system. Devanagari script is a two dimensional form of symbol. It is very inconvenient to treat each form of character as a separate symbol because such combinations are very large in number. Segmentation is a process which separates words into characters. In this paper, we deal with segmentation of offline handwritten document especially touching character segmentation also we have discussed the process involved in touching character segmentation and the problem in those processes with their solution. The proposed method is tested on the database set of 250 handwritten touching characters. Experimental results indicate that the method used correctly segments the touching characters.

Handwritten Devanagari Character Classification Using CNN

Abstract: Deep Convolutional Neural Network (CNN) has been effectively outlined acknowledgement of the written by hand character in Devanagari content. Convolutional Neural Networks build up a framework that can be set up to perceive the morphology of pictures and perform class estimate and concentrate the property of pictures. Written by hand character acknowledgement is a huge undertaking as it tries to perceive the right class for client autonomous manually written digits. Preparing of the model is finished utilizing MNIST dataset This challenge is much more difficult with a heavily impressionistic, morphologically complicated, and inherently juxtapositional character consisting of dialect such as Devanagari. A dataset of 73,600 samples were trained and tested in this proposed CNN that are separated to perform both training and testing which provided quite impressive results. 97% exactness was gotten for the test information when the system was tried with a Devanagari character dataset.

Script Identification of Handwritten Images

Abstract: This paper describes a system for script identification of handwritten word images. The system is divided into two main phases, training and testing. The training phase performs a moment based feature extraction on the training word images and generates their corresponding feature vectors. The testing phase extracts moment features from a test word image and classifies it into one of the candidate script classes using information from the trained feature vectors. Experiments are reported on handwritten word images from three scripts: Latin, Devanagari and Arabic. Three different classifiers are evaluated over a dataset consisting of 12000 word images in training set and 7942 word images in testing set. Results show significant strength in the approach with all the classifiers having a consistent accuracy of over 97%.