4 A Comparative Analysis of Machine Learning Techniques for Odia Character Recognition
-
Mamatarani Das
Abstract
Character recognition is a challenging area in Machine Learning, Pattern Recognition or Image Processing. The accuracy to recognize handwritten character by human is far better compared to machine recognition. To develop an interface which can differentiate characters written by human yet requires intensive research. Though number of researches have presented in this area, still research is going on to achieve human like accuracy. Both handwritten and printed character recognition are categorized into two types, online and offline. A good number of researches have done work in the area of optical character recognition in different languages but for the Odia language, development is negligible. Odia (formerly it was Oriya), one of the 22 scheduled language recognized by the constitution of India and it is the official language of the state of Odisha (Orissa), more than 40 million people speak Odia. Due to the roundish shape of Odia character, large number of modified and compound characters and similarity between different characters makes this language very hard to create a satisfactory classifier. In the present survey undertaken we have discussed what are challenges be for Odia language and the machine learning techniques used in the recognition of Odia character recognition. This chapter describes complete process of character recognition i. e. pre-processing, extraction and selection of feature set and character recognition elaborately with comparison analysis and the metrics used to evaluate machine learning algorithms.
Abstract
Character recognition is a challenging area in Machine Learning, Pattern Recognition or Image Processing. The accuracy to recognize handwritten character by human is far better compared to machine recognition. To develop an interface which can differentiate characters written by human yet requires intensive research. Though number of researches have presented in this area, still research is going on to achieve human like accuracy. Both handwritten and printed character recognition are categorized into two types, online and offline. A good number of researches have done work in the area of optical character recognition in different languages but for the Odia language, development is negligible. Odia (formerly it was Oriya), one of the 22 scheduled language recognized by the constitution of India and it is the official language of the state of Odisha (Orissa), more than 40 million people speak Odia. Due to the roundish shape of Odia character, large number of modified and compound characters and similarity between different characters makes this language very hard to create a satisfactory classifier. In the present survey undertaken we have discussed what are challenges be for Odia language and the machine learning techniques used in the recognition of Odia character recognition. This chapter describes complete process of character recognition i. e. pre-processing, extraction and selection of feature set and character recognition elaborately with comparison analysis and the metrics used to evaluate machine learning algorithms.
Chapters in this book
- Frontmatter I
- Dedication V
- Preface VII
- Inhalt XI
- 1 A prologue to Emerging trends in Machine Learning 1
- 2 An Analysis on Non-linear Dimension Reduction Techniques 19
- 3 Application of Machine Learning in Music Analytics 43
- 4 A Comparative Analysis of Machine Learning Techniques for Odia Character Recognition 65
- 5 Pre Filtering with Rule Mining for Context Based Recommendation System 91
- 6 Early detection of crop diseases using machine Learning based intelligent techniques: a review 115
- 7 Conclusion 131
- Contributing authors 137
- Index 139
Chapters in this book
- Frontmatter I
- Dedication V
- Preface VII
- Inhalt XI
- 1 A prologue to Emerging trends in Machine Learning 1
- 2 An Analysis on Non-linear Dimension Reduction Techniques 19
- 3 Application of Machine Learning in Music Analytics 43
- 4 A Comparative Analysis of Machine Learning Techniques for Odia Character Recognition 65
- 5 Pre Filtering with Rule Mining for Context Based Recommendation System 91
- 6 Early detection of crop diseases using machine Learning based intelligent techniques: a review 115
- 7 Conclusion 131
- Contributing authors 137
- Index 139
