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.

Joint Transform Correlator Applied to Recognition of Devanagari Script

Abstract: A hybrid optoelectronic character recognition system is proposed to identify Devanagari scripts which is also used in old Sanskrit manuscripts. In this system, the correlation operation is performed using a joint-transform correlator (JTC) and data-transfer is done using electronic peripheral hardware. Optical system design of JTC using a liquid crystal spatial light modulator and CCD (charge coupled device) camera was performed. Using a one-channel JTC, we demonstrated for the first time the discrimination of all core elements. Furthermore, novel methods using special characteristics of the script such as syllable partitioning and skew angle correction are discussed.

Handwritten Devanagari Character Generation Using Deep Convolutional Generative Adversarial Network

Abstract: As deep learning became popular, the need for huge amounts of data has risen. The major problem faced in deep learning is the data scarcity. Many researchers have done research in areas such as image processing, pattern recognition, artificial intelligence, and cognitive science to solve handwritten character recognition problem but the data availability remains the problem particularly in Indian languages. The main motive of this paper is to generate the handwritten character of Devanagari, for which DCGANs are used which help us to generate training data images from the vector representation. Here, we use three-layer CNN having a stride value of two for feature extraction of the handwritten character. The characters generated look like the character in the original dataset.

Saraswati: cross‐lingual Sanskrit Digital Library

Abstract: Purpose – The purpose of this paper is to describe Saraswati, a cross‐lingual Sanskrit Digital Library hosted at Banaras Hindu University. The system aims to assist those who know Sanskrit and at least one Indic script out of Devanagari, Kannada, Telugu and Bengali.Design/methodology/approach – The system is developed with the Unicode standard using PHP as the programming language. The system follows three levels of architecture for search, display, and storage of Sanskrit documents. The system uses the UTF‐8 character representation system and generates on‐the‐fly transliteration from one Indic language script to another.Findings – The system successfully demonstrates transliteration of Sanskrit text from one language to another. Saraswati is also capable of searching a given keyword across different languages and produces the result in the desired language script.Research limitations/implications – Some languages such as Tamil (not chosen for study) use context dependent consonants, and with the present...

Handwritten Devanagari Numeral Recognition by Fusion of Classifiers

Abstract: Handwritten Devanagari Numeral Recognition by Fusion of Classifiers Recognition of handwritten Devanagari numerals has many applications especially in the field of postal automation, document processing and so on. Due to its vast applications, many researchers are actively working towards development of effective and efficient hand written character/numeral recognition. Devanagari script is widely used script in Indian sub-continent; also Devanagari script forms the basis for many other scripts in Indian sub-continent. In this paper, we have proposed a hybrid method to recognize handwritten Devanagari numerals. The proposed method uses, stacking approach to fuse the confidence scores from four different classifiers viz., Naive Bayes (NB), Instance Based Learner (IBK), Random Forest (RF), Sequential Minimal Optimization (SMO). Also, the proposed method extracts both local and global features from the handwritten numerals. In this work, we have used Fourier Descriptors as global shape feature. Whereas, the pixel density statistics from different zones of the numeral to describe the numerals locally. The proposed method has been tested on large set of handwritten numeral database and experimental results reveal that the proposed method yields the accuracy of 99.685%, which is the best accuracy reported so far for the datasets considered. Hence the proposed method outperforms contemporary algorithms.