CHAPTER ONE

INTRODUCTION

1.1     Background of the Study

Handwritten Character Recognition (HCR) is the ability of a computer to recognize the human handwritten digits from different sources like images, papers, touch screens, etc, and classify them into 10 predefined classes (0-9). This has been a topic of boundless-research in the field of deep learning. Digit recognition has many applications like number plate recognition, postal mail sorting, bank check processing, etc. In Handwritten digit recognition, everyone faces many challenges because of different styles of writing of different peoples as it is not an Optical character recognition system (Dixit et al., 2020).

Handwritten Character Recognition is a field of research in image contrast, computer vision and style recognition. An average computer will gain the ability to make a distinction between the characters on photographs, paper records, touch screen gadgets from different resources and then convert into machine-readable notations. This helps in a wide variety of computer applications dealing with optical character reception and framework development for character handling. One of the notable issues with HCR is Picture Rating. It is the process of sending input pictures to analyze from an invariable arrangement of surrounding reliant pictures. In Optical Character Recognition (OCR), calculations are conducted on a data collection consisting of recognized characters and interprets how to group these characters that are found from the test set (Budhi et al., 2015).

Handwritten Character Recognition (HCR) has been the most complicated and intriguing research area in the areas of digital image processing and pattern recognition. An accurate system of recognition makes major contribution in making improvement in intelligent and automatic systems to enhance the interface between humans and technology in various applications. The major focus of the modern research is to develop algorithms that minimize the processing time while maintaining high accuracy rates (Adeel et al., 2019).

Broadly, HCR systems are divided into two categories: online and offline. The online recognition systems are dependent on two of the coordinates expressed as the function of number and time and also the strokes’ order that writer made. Special hardware like a tablet PC and a pen is required in case. Experiments have proven that some of the online procedures are more accurate when referring to character recognition when compared with other offline methods because of additional time dependent information that is seen in the other method. In offline recognition, the handwriting is accessible as an image. Offline recognition is attained through different applications like processing of bank cheque, handwritten form processing and document reading etc. That is why, offline systems are the major areas of research that have been used for making improvement in rates of recognition (lamondon and Srihari, 2011).

In recent times, more research is being carried out to design a robust system for HCR and various methodologies are being explored to find out one that gives desired results for some particular application. For example, Artificial Neural Networks have now become more effective in making improvement in the accuracy of character recognition in offline systems because of accuracy, speed and robustness (Khan et al., 2018). The proposed system aims to use convolution neural network, a class of deep learning algorithm to recognized handwritten character digit.

1.2     Statement of the Problem

The recognition of handwritten digit, that has been poorly written, has been a challenge in several areas such as such as offices, schools, hospitals, banks, and other places where handwritten documents are handled more frequently. This problem has slowed down many processes in this areas, as a handwritten digit that is poorly written would take longer time, and as well as strain the eyes of the individual trying to comprehend what has been poorly written. Many aspects of human experience are been automated for improved work efficiency and experience, taking into consideration the far increasing need for automation in various sector, a deep learning-based approach to handwritten digit recognition is been proposed in order to reduce the laborious process of human eye recognition of handwritten digits that are poorly written.

1.3     Aim and Objectives of the Study

The aim of this research is to develop a handwritten digit character recognition system that will use a deep learning (DL) algorithm known as Convolutional Neural Network (CNN) to recognize handwritten digits.

The objectives of the study are as follows to:

1. Development of a deep learning model for recognizing handwritten/character digits.
2. Reduce the time frame it takes the human eye in the recognition of various human hand writing digits due to different writing styles of people, by deploying a trained model to carry out the recognition task.
3. Test the deployed trained model with photos that were not included in the model's training dataset to access the developed system's performance.

1.4     Significance of the Study

The system, handwritten digit character recognition system will be of significant benefits as thus;

1. The corporate sector, via the deployment of the developed software along corporate professionals in charge of duties and responsibility examining characters digits in their line of duty, which will in turn bring about more ease and less time in the recognition of characters.
2. Supermarket stores for the recognition of price digits on products.

1.5     Scope of the Study

The system proposed in this research will cover the recognition of characters. The system would take in input image of characters and make prediction based on the model being trained and deployed using convolutional neural network (CNN) algorithm on the system.

1.6     Limitation of the Study

The researcher faced some limitations during the course of this research. Time factor was the major factor as the timeframe for the proposed system was too short for extensive research. Lack of adequate finance to fund the researcher is also a constraint, as to outsource many data.

1.7     Definition of Terms

1. Artificial Intelligence (AI): This is intelligence demonstrated by machines, unlike the natural intelligence displayed by humans and animals, which involves consciousness and emotionality.
2. Digit: A digit is an element of a set that, taken as a whole, comprises a system of numeration. Thus, a digit is a number in a specific context. In the decimal (base-10) Arabic numbering system, the digits are the elements of the set {0, 1, 2, 3, 4, 5, 6, 7, 8, 9}.
3. Handwritten Digit Recognition: Handwritten digit recognition is the ability of a computer to recognize the human handwritten digits from different sources like images, papers, touch screens, etc, and classify them into 10 predefined classes (0-9).
4. Handwritten: Handwritten and hand-written are both adjectives which mean written using a pen or pencil, instead of being printed or typed.
5. Convolutional Neural Network (CNN): This is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various aspects/objects in the image and be able to differentiate one from the other.
6. Support Vector Machine (SVM): is a machine learning algorithm, support-vector machines are supervised learning models with associated learning algorithms that analyze data for classification and regression analysis.
7. Data Argumentation:   Is use in deep learning to regularize, normalize the information on data to be trained.
8. Data Notation:    This is the process of labeling data to make it usable for machine learning.
9. Dataset: A collection of related sets of information that is composed of separate elements but can be manipulated as a unit by a computer.
10. Deep Learning (DL): This is a subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks.
11. Machine Learning (ML): This is the study of computer algorithms that improve automatically through experience and by the use of data.
12. Mobile App:  is a computer program or software application designed to run on a mobile device such as a phone-tablet or watch.
13. Model: is a description or analogy used to help visualize something that cannot be directly observed.
14. Neural Networks (NN): This is a series of algorithms that endeavors to recognize underlying relationships in a set of data through a process that mimics the way the human brain operates.
15. Perceptron: This refers to a linear classifier (binary) use in supervised learning. It helps to classify the given input data.
16. Software:   It is a program and other operating information use by a computer system.
17. Tensor-Flow: is a free and open-source software library for dataflow and differentiable programming across a range of tasks. It is a symbolic math library, and is also used for machine learning applications such as neural networks.
18.  Visual Geometry Group (VGG): is a machine learning model, whose goal of the Visual Geometry Group (VGG) is to build this sense of vision in machine and artificial intelligence.
19. Multi-layer Perceptron (MLP): A multilayer perceptron is a fully connected class of feedforward artificial neural network which consists of three types of layers, the input layer, output layer and hidden layer.