Handwritten Character Recognition Using Deep Learning With Implementation (Matlab)

Electrical Electronics Engineering

Handwritten character recognition, a crucial aspect of optical character recognition (OCR), employs deep learning methodologies for efficient recognition of handwritten characters. Leveraging neural networks, particularly convolutional neural networks (CNNs), this approach involves preprocessing steps such as image resizing, normalization, and noise reduction to enhance the quality of input data. Subsequently, the CNN architecture, comprising convolutional layers, pooling layers, and fully connected layers, learns intricate patterns and features from the input images. Implementation in MATLAB involves dataset preparation, model training, validation, and testing phases. Utilizing frameworks like MATLAB’s Deep Learning Toolbox facilitates the construction and training of CNN models, while techniques like transfer learning can expedite the process by leveraging pre-trained networks. Moreover, fine-tuning hyperparameters, such as learning rates and batch sizes, optimizes model performance. Post-training evaluation involves assessing metrics like accuracy, precision, and recall to gauge the model’s efficacy in recognizing handwritten characters. Additionally, integrating techniques like data augmentation can enhance the model’s robustness and generalization capabilities, especially when faced with variations in handwriting styles and writing conditions. Overall, leveraging deep learning techniques in MATLAB empowers accurate and efficient handwritten character recognition, with potential applications in document digitization, text analysis, and automated data entry.

ABSTRACT

To identify a handwritten character is always a challenging task in the field of research on image processing, artificial intelligence as well as machine vision since the handwriting varies from person to person. Moreover, the handwriting styles, sizes and its orientation make it even more complex to interpret the text. The numerous applications of handwritten text in reading bank cheques, Zip Code recognition and in removing the problem of handling documents manually has made it necessary to acquire digitally formatted data. This work is on the recognition of handwritten characters using Matlab, followed by the implementation of various other Matlab toolboxes like Image Processing and Neural Network Toolbox to process the scanned or acquired image. Experimental Results are given to present the proposed model in order to recognize handwritten characters accurate

TABLE OF CONTENTS

COVER PAGE

TITLE PAGE

APPROVAL PAGE

DEDICATION

ACKNOWELDGEMENT

ABSTRACT

CHAPTER ONE

INTRODUCTION

BACKGROUND OF THE STUDY
PROBLEM STATEMENT
AIM/OBJECTIVE OF THE STUDY
PURPOSE OF THE STUDY
RESEARCH QUESTION
SIGNIFICANCE OF THE STUDY
SCOPE OF THE STUDY
LIMITATION OF THE STUDY
RESEARCH METHODOLOGY
PROJECT ORGANISATION

CHAPTER TWO

LITERATURE REVIEW

REVIEW OF THE STUDY
OVERVIEW OF MACHINE LANGUAGE
HISTORICAL BACKGROUND OF MACHINE LANGUAGE
REVIEW OF RELATED STUDIES
CLASSIFICATION OF HANDWRITING CHARACTER RECOGNITION

CHAPTER THREE

METHODOLOGY

WORKING PRINCIPLE
METHOD USED

CHAPTER FOUR

RESULT ANALYSIS

RESULTS

CHAPTER FIVE

CONCLUSION
REFERENCES

CHAPTER ONE

1.0 INTRODUCTION

1.1 BACKGROUND OF THE STUDY

With the advancement of technology, the interfacing between man and machine has increased the scope of research in various domains, thereby making majority of tasks automated and easier to perform.

The development of handwriting recognition systems began in the 1950s when there were human operators whose job was to convert data from various documents into electronic format, making the process quite long and often affected by errors. Automatic text recognition aims at limiting these errors by using image preprocessing techniques that bring increased speed and precision to the entire recognition process. Handwriting recognition has been one of the most fascinating and challenging research areas in field of image processing and pattern recognition in the recent years. It contributes immensely to the advancement of automation process and improves the interface between man and machine in numerous applications.

Over years, organizations get on hold to numerous numbers of handwritten documents, forms and checks. Preserving such paper documents is a very tedious and a time-consuming task. Also over the years these handwritten paper documents might get distorted, and no longer be of use. In order to preserve such handwritten paper documents, if converted to a digital format it would reduce retrieval process and also make handling of such documents easier and reliable.

Optical Character Recognition has been a widely important research topic in pattern recognition, machine vision and artificial learning. Optical Character Recognition is also known as off-line character recognition system.

Optical Character Recognition at times fails to recognize the handwritten text as the writing style varies from person to person. The main task of an Optical Character Recognition is to successfully identify printed text and recognize it. Optical Character Recognition includes the following phases namely, Pre- processing, Segmentation, Feature Extraction and Classification with Recognition. Output of one-phase acts as an input of the next phase in process.

Various novels written by authors are handwritten, these are edited first and then typed to generate a final copy and then printed to produce a book. If a suitable Handwritten Character Recognition System (HCRS) is made available, then these difficulties can be overcome. A HCRS with a high accuracy rate, which can convert any handwritten document to a digital image format, would overcome all anomalies. In some handwritten documents, a common problem faced is to successfully be able to distinguish each and every character successfully. As the writing style of very individual varies, to identity and distinguish them is difficult even for the human eye. At times the same person might not be able to identify the character written by him/her, also he/she might not always write the same character in a similar manner.

The main outcome of a HCRS would be to identify and recognize different handwriting styles. There are various number of ways to write a single character, which differs individual to individual.

This work is on Handwritten Character Recognition using Neural Network in Matlab. MATLAB is one such powerful machine tool where in the availability of Image Acquisition Toolbox, Image Processing Toolbox and Neural Network Toolbox simplifies the task of obtaining and understanding handwritten text.

Hence, the aim is to achieve a high accuracy rate by classifying and recognizing characters to the extent feasible.

1.2 PROBLEM STATEMENT

Handwritten character recognition is a complex problem, which is not easily solvable due to the fact that handwriting varies from person to person. To solve the defined handwritten character recognition problem of classification we used MATLAB computation software with Neural Network Toolbox and Image Processing Toolbox add-on. The purpose is to develop Handwriting Character Recognition Software with a higher accuracy rate reducing its space and time complexities; making it optimal.

AIM AND OBJECTIVE OF THE STUDY

The main aim of this work is to carry out a study on handwritten character recognition using convolutional neural networks (CNNs) machine language which is implemented using MATLAB. The objectives are:

The objective of this study is to identify handwritten characters with the use of neural networks
To use deep learning toolbox to carry out a simulation studies using MATLAB for performance

To reduce retrieval process and also make handling of documents easier and reliable.

1.4 PURPOSE OF THE STUDY

The purpose is to develop Handwriting Character Recognition Software with a higher accuracy rate reducing its space and time complexities; making it optimal

RESEARCH QUESTION

What are the advantages of hand written recognition?
How does handwriting recognition work?

How can you identify your handwriting using deep learning?

1.6 SIGNIFICANCE OF THE STUDY

The role of Handwriting recognition has led to carrying out this study. Handwriting recognition plays a big role in the technology world now. It also plays an important role in the storage and in the recovery of critical handwriting information. This handwriting recognition ensures an accurate medical care and it also reduces storage costs. It ensures that an essential field of research remains available to students in the future. In this era of globalization, technologies continue to improve and improve more in no time.

1.7 SCOPE OF THE STUDY

Using Matlab Neural Network toolbox, we tried to recognize handwritten characters by projecting them on different sized grids. The first step is image acquisition which acquires the scanned image followed by noise filtering, smoothing and normalization of scanned image, rendering image suitable for segmentation where image is decomposed into sub images. Feature Extraction improves recognition rate and classification. We use character extraction and edge detection algorithm for training the neural network to classify and recognize the handwritten characters.

LIMITATION OF THE STUDY

As we all know that no human effort to achieve a set of goals goes without difficulties, certain constraints were encountered in the course of carrying out this project and they are as follows:-

Difficulty in information collection: I found it too difficult in laying hands of useful information regarding this work and this course me to visit different libraries and internet for solution.
Financial Constraint: Insufficient fund tends to impede the efficiency of the researcher in sourcing for the relevant materials, literature or information and in the process of data collection (internet, questionnaire and interview).

Time Constraint: The researcher will simultaneously engage in this study with other academic work. This consequently will cut down on the time devoted for the research work.

1.9 RESEARCH METHODOLOGY

In the course of carrying this study, numerous sources were used which most of them are by visiting libraries, consulting journal and news papers and online research which Google was the major source that was used.

1.10 PROJECT ORGANISATION

The work is organized as follows: chapter one discuses the introductory part of the work, chapter two presents the literature review of the study, chapter three describes the methods applied, chapter four discusses the results of the work, chapter five summarizes the research outcomes and the recommendations.