A Literature Review on Hand-written Character Recognition with Neural Network

Abstract

The handwriting styles of the various individuals are unique in this manner; it is truly challenging to recognize the manually written characters. Manually written character acknowledgment is a space of example that has turned into the subject of exploration during the last few decades. The neural organization assumes a significant part in transcribed person acknowledgment. This requires the study and reproduction of Artificial Neural Networks. In Neural Network, every hub plays out a few straightforward calculations and every association passes a sign starting with one hub on then onto the next marked by a number called the \"association strength\" or weight, demonstrating the degree to which sign is intensified or reduced by the association. Hindi content is written from left to right and from start to finish design.

Introduction

I. INTRODUCTION

According to many reports, Hindi is acknowledged as being written by many. There is still a chance for better accuracy and a lack of auto coding with neural networks to work on the most commonly used characters in the language. As such, there is an open need for a person-written methodology to learn and teach Hindi so that all recognition will be accurate.

The ability of the PC to get, and decipher clear transcribed contributions from sources, for example, paper documents, photographs, touch screens, and other devices. This will be accomplished by displaying a neural organization that should be prepared over a dataset. Handwriting acknowledgment frameworks use design coordination to change over manually written letters into comparing PC messages or orders continuously.

This paper is directed toward creating programming that will be able to understand and process characters in the Hindi language. What neural networks are especially helpful for is taking care of issues that can't be communicated as a progression of steps; for example, perceiving designs, ordering them into gatherings, series expectations, and information mining. Artificial Neural Networks also perceive different articles better than humans do and seem to understand with ease despite the huge volume of visual data, which requires hardly any work from them. It is the goal of this paper to recreate the errands performed by Neural Networks to see how close it can get to how a human understands things by restricting its abilities with constraints such as limited memory size and processing power.

II. RELATED WORK

R. Vijaya Kumar Reddy, U. Ravi Babu [1] proposed  a manually written Hindi person recognition framework in light of various Deep learning procedures. Manually written character recognition is taking up a significant role and is now being noticed by experts given potential applications in helping innovation for blind and disabled clients. Here, the Convolutional Neural Network (CNN) and Deep Feed Forward Neural Networks (DFFNN) are trained and tested on a standard client dataset that is gathered from different clients. From the test results, DFFNN, CNN gives the best exactness for Handwritten Hindi characters when contrasted to the alternative procedures. Here the proposed technique produces pleasing outcomes with a high accuracy rate.

Ajay Indian, Karamjit Bhatia [2] proposed a broad outline of examination work, which have been finished for perceiving the disconnected manually written characters utilizing the different methodologies like-ANNs, Fuzzy Logic, Genetic calculations, SVM, KNN Hidden Markov Model (HMM), Bacterial Foraging, Clonal Selection Calculation and so forth and talk about their exhibition as far as exactness of perceiving the characters. It has been seen that precision of Offline Handwritten Hindi Character Recognition relies upon division process, neighborhood as well as worldwide elements separated; an assortment of classifiers frameworks and mix of techniques took on. There is a 100% of the time extent of progress in the precision of Offline Handwritten Hindi Character Recognition (OFHCR) by utilizing different effective nearby elements, worldwide elements extraction.

Vinod Jha, K. Parvathi,[3] proposed Braille is a language of most extreme significance for the upliftment of visually impaired individuals. One of the issues with schooling of visually impaired individuals is the inaccessibility of assets, for example, books, and so on, which are for the most part in nearby dialects being utilized by broad individuals. As of late there have been heaps of works done to mechanize the method involved with interpreting books in English, Arabic, and so on. Accuracy level of this 95.56% as compared before.

Muna Ahmed Awel, Ali Imam Abidi [4] Their work analyzes character recognition in English, Arabic, and Devanagiri characters. They clarify the techniques they use and the challenges they face during the improvement of optical person acknowledgment. In Optical Scanning, Pre-handling, Segmentation, Feature extraction, and so on, there are major steps involved with character recognition. The examination work reconsidered in this paper utilizes various methodologies to achieve great accuracy.

R. Vijaya Kumar Reddy, Dr. B. Srinivasa Rao [5] proposed a handwritten Hindi character recognition, a novel approach has been proposed. The most challenging aspect of identifying these characters is that journalists have different penmanship styles in every dialect. For these types of characters, a neural network achieved an accuracy rate of 99.85% with the proposed strategy.

Prasanta Pratim Bairagi [7] proposed a Optical Character Recognition is the mechanical or electronic interpretation of pictures of transcribed or printed text into machine-editable text. Over the last 50 years, this method has proven to be a powerful way of capturing information. It is quicker, more accurate, and a better alternative to typing. The accuracy depends on how many different images are scanned.

III. PROPOSED SYSTEM

Neural networks offer total autonomy in acknowledgment interaction and character set. In this neural network is first prepared by different example pictures of every letters in order. Then, in the acknowledgment process, the input picture is straightforwardly given to neural organization and a perceived image is yielded. Deep learning is a broad branch of artificial intelligence that is human-mimicking. Unlike other machine learning methods, deep learning does not require explicit programming; instead, it relies on neural networks to mimic the human brain. Deep learning has been around for years, but it is more popular now because of the new computational power and larger data sets.

Digitization

Pre-processing.

Feature Extraction.

Classification and Recognition

IV. DISCUSSION

The neural network is given an objective vector and furthermore a vector that contains the example data; this could be a picture and written by hand information. The neural network then, at that point, endeavors to decide whether the information matches an example that the brain network has remembered. A neural network prepared for characterization is intended to take input tests and arrange them into gatherings. These gatherings might be fluffy, without plainly characterized limits. This paper concerns recognizing free written by hand characters. The reason for this task is to accept, written by hand Hindi characters as information, process the character, train the neural network algorithm, perceive the example, and adjust the system to an embellished adaptation of the input. This paper is directed towards creating programming which will be supportive in perceiving characters of the Hindi language. One of the essential means by which PCs are supplied with human-like capacities is using neural organization. Neural networks are especially helpful for tackling issues that can't be communicated as a progression of steps; for example, perceiving patterns, classifying them into groups, series expectations, and information mining.

Conclusion

This article alluded to a framework for Hindi content. Since this is still a developing capability, the accuracy of recognition is promising, but more research needs to be done. This research should include pictures with Hindi text in various fonts and sizes. After enough exploration, I have found that handwriting recognition has been one of the most compelling and challenging research areas in the fields of computer vision and artificial intelligence.

References

[1] R. Vijaya Kumar Reddy, U. Ravi Babu (Feb 2019). Handwritten Hindi Character Recognition using Deep Learning Techniques. In 2019 International Journal of Computer Sciences and Engineering (IJCSE). (Vol.-7, Issue-2,). [2] Ajay Indian, Karamjit Bhatia (2017) A Survey of Offline Handwritten Hindi Character Recognition. In 2019 International Conference on Inclusive Technologies and Education (CONTIE). IEEE. [3] Vinod Jha, K. Parvathi,(2019) Braille Transliteration Of Hindi Handwritten Texts Using Machine Learning For Character Recognition. In 2019 INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH (VOLUME 8, ISSUE 10). IJSTR. [4] Muna Ahmed Awel, Ali Imam Abidi (2019) REVIEW ON OPTICAL CHARACTER RECOGNITION. In 2019 International Research Journal of Engineering and Technology (IRJET). (Vol-06, Issue-06). [5] R. Vijaya Kumar Reddy, Dr. B. Srinivasa Rao, K. Prudvi Raju(2018).Handwritten Hindi Digits Recognition Using Convolutional Neural Network with RMSprop Optimization. In 2018 Proceedings of the Second International Conference on Intelligent Computing and Control Systems (ICICCS). IEEE . [6] Nicole Dalia Cilia (2018). A ranking-based feature selection approach for handwritten character recognition. In 2018 Accelerating the world\'s research. (ELSEVIER). (www.elsevier.com/locate/patrec).(Vol-121). [7] Prasanta Pratim Bairagi (2018) Optical Character Recognition for Hindi. In 2018 International Research Journal of Engineering and Technology (IRJET) (Volume: 05 Issue: 05). [8] Abhishek Mehta, Dr. Subhashchdra Desai ,Dr. Ashish Chaturvedi (2021).Hindi Handwritten Character Recognition from Digital Image using Deep Learning Neural Network. In 2021 International Journal of Engineering Research & Technology (IJERT).( Volume 9, Issue 5). [9] Rumman Rashid Chowdhury, Mohammad Shahadat Hossain,(2019).Bangla Handwritten Character Recognition using Convolutional Neural Network with Data Augmentation. In 2019 8th International Conference on Informatics, Electronics & Vision (ICIEV). [10] Vinoj Jayasundara, Sandaru Jayasekara, Hirunima Jayasekara (2019). Handwritten Character Recognition with Very Small Datasets. In 2019 Winter Conference on Applications of Computer Vision (WACV) IEEE.

Copyright

Copyright © 2022 Priti Singh Rathore, Dr. Pawan Kumar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.