A Novel Approach Towards Handwritten Digit Recognition Using Refraction Property of Light Rays

A Novel Approach Towards Handwritten Digit Recognition Using Refraction Property of Light Rays

Roopkatha Samanta (Jadavpur University, India), Soulib Ghosh (Jadavpur University, India), Agneet Chatterjee (Jadavpur University, India) and Ram Sarkar (Jadavpur University, India)
Copyright: © 2020 |Pages: 17
DOI: 10.4018/IJCVIP.2020070101
OnDemand PDF Download:
No Current Special Offers


Due to the enormous application, handwritten digit recognition (HDR) has become an extremely important domain in optical character recognition (OCR)-related research. The predominant challenges faced in this domain include different photometric inconsistencies together with computational complexity. In this paper, the authors proposed a language invariant shape-based feature descriptor using the refraction property of light rays. It is to be noted that the proposed approach is novel as an adaptation of refraction property is completely new in this domain. The proposed method is assessed using five datasets of five different languages. Among the five datasets, four are offline (written Devanagari, Bangla, Arabic, and Telugu) and one is online (written in Assamese) handwritten digit datasets. The approach provides admirable outcomes for online digits whereas; it yields satisfactory results for offline handwritten digits. The method gives good result for both online and offline handwritten digits, which proves its robustness. It is also computationally less expensive compared to other state-of-the-art methods including deep learning-based models.
Article Preview

1. Introduction

Handwritten digit recognition (HDR) is the ability of a computer to elucidate an obtained comprehensible handwritten input from sources such as paper documents, images and some other devices. Offline HDR implies transformation of the input into some other form which can be used by computer and other applications necessary for the recognition procedure. The data obtained in this process is regarded as a static representation of handwritten digit. Though machine printed digit and character recognition problems are almost sorted, handwritten digit and character recognition still need much effort to accomplish. So, HDR is such a field of research which needs more cultivation. In addition to that, in India, different languages are used in different states. For that we need a general and relatively simple approach that can handle the digit images written in multiple languages with ease. HDR is considered as an important module of an Optical Character Recognition (OCR) system to be developed for recognizing handwritten textual document images. This involves photo-scanning of the text taking one character at a time, analysis of the scanned-in image, and then conversion of the scanned-in image into character codes, such as ASCII, generally used in data processing through computer system. HDR is also used in: (a) national ID number recognition system (b) postal office automation with code number recognition on Envelope and (c) handwritten cheque processing. Recently HDR has been implemented in tablet-based learning and e-content development. The challenges in handwritten digit recognition arise not only from the different ways in which a single digit can be written, but also from the different requirements imposed by some specific applications. Writing style varies from person to person and even it differs for a single person. Thus, it is a more difficult issue to recognize unconstrained handwritten digits. Any prior hint is also not provided from which one can get any idea about the writing style. So this is a very challenging and practical problem to manage. This implies that the main demand of the system is to develop such a method which will be able to detect the handwritten digit images irrespective of its writing style. Apart from these, some key challenges also includes uneven stroke width, irregular illumination, non-uniform orientation, seepage of ink etc. Some of the difficulties faced in HDR are shown in Table 1.

Table 1.
Example of challenges faced during recognition of handwritten digits. The images are taken from handwritten Bangla digit dataset. Figure (a), (b), (c), (d) are examples of poor handwriting, seepage of ink, uneven stroke width, and non-uniform orientation respectively.

Complete Article List

Search this Journal:
Volume 13: 1 Issue (2023): Forthcoming, Available for Pre-Order
Volume 12: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 11: 4 Issues (2021)
Volume 10: 4 Issues (2020)
Volume 9: 4 Issues (2019)
Volume 8: 4 Issues (2018)
Volume 7: 4 Issues (2017)
Volume 6: 2 Issues (2016)
Volume 5: 2 Issues (2015)
Volume 4: 2 Issues (2014)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing