Addis Ababa University School of Graduate Studies IMAGE ANALYSIS FOR ETHIOPIAN COFFEE CLASSIFICATION By Habtamu Minassie Aycheh A Thesis Submitted to the School of Graduate Studies of Addis Ababa University in Partial Fulfillment for the Degree of Master of Science in Computer Science January 2008 ADDIS ABABA UNIVERSITY SCHOOL OF GRADUATE STUDIES FACULTY OF INFORMATICS DEPARTMENT OF COMPUTER SCIENCe IMAGE ANALYSIS FOR ETHIOPIAN COFFEE CLASSIFICATION By Habtamu Minassie Aycheh MEMBERS OF THE EXAMINATION BOARD: 1. Sebsibe Hailemariam, Advisor ______________________ 2. _________________________ ______________________ 3. _________________________ _______________________ DEDICATION To GOD For Blessing Ethiopia with Natural Coffee - Our Green Gold To the Memory of My Brother Yizengaw Makonnen For His Great Inspirations in My Early Ages To My Parents and My Beloved Fiance – Bezuyie ACKNOWLEDGEMENTS First of all, I would like to express my deepest gratitude to my advisor Sebsibe Hailemariam for his motivative and constructive guidance right from the moments of problem formulation to the completion of the work. Many thanks and appreciations go to him for the discussions with him always made me think that things are possible. His enthusiasm and encouragement has always inspired me to accelerate to the completion of the work. I am also very thankful to my instructors and all staffs members of the department of Computer Science for their contribution in one way or another for the success of my study. I would like to forward my special thanks to the Ethiopian Coffee Quality Inspection and Auction Center for supplying samples of coffee varieties. I would like to extend my appreciation and thanks to laboratory experts and all staffs members of the organization for their kind cooperation. Particularly, many thanks go to Ato Endale Asfaw for his constant assistance in explaining the work in the problem domain and giving me reading materials. I would also like to thank Rasband Wayne, the developer of ImageJ software for free availability of the resource above all and for giving prompt reply to my questions and helped a lot in customizing plugins. In addition, I would like to thank my friend Hassen Redwan for his tireless support in printing the materials. The many times discussions and sharing of ideas and resources with him had a significant contribution to the success of this work. Most of all, I wish to thank my beloved fiance, Bezunesh Tesfaye, for her caring in all my ways. I am also very thankful to my brother, Abebe Minassie, for encouraging and supporting me in all my studies starting from early schools. I extend my wish to thank my long-term friends Terefe Niguise, Dawit Bulcha and Atnaw Wubshet for their valuable supports and standing by me in all the difficult moments. Finally, I am very grateful to my father and mother and all the rest of my families, friends, and peers who, in one or the other way brought me up to a success in my academic endeavor. i TABLE OF CONTENTS ACKNOWLEDGEMENTS ......................................................................................................i LIST OF TABLES...................................................................................................................iv LIST OF FIGURES..................................................................................................................v ABSTRACT............................................................................................................................ vii 1. INTRODUCTION ............................................................................................................1 1.1 Background ...............................................................................................................1 1.2 Motivation..................................................................................................................4 1.3 Problem Statement ...................................................................................................5 1.4 Scope of the Research...............................................................................................6 1.5 Objective....................................................................................................................7 1.6 Methodology..............................................................................................................7 1.6.1 Materials and methods............................................................................................8 1.6.2 Sampling Techniques..............................................................................................8 1.6.3 Tools .......................................................................................................................9 1.6.4 Classification Approaches ....................................................................................10 1.7 Layout of the Thesis................................................................................................10 2. LITERATURE REVIEW ..............................................................................................12 2.1 Ethiopian Coffee .....................................................................................................12 2.1.1 Coffee Processing .................................................................................................13 2.1.2 Coffee Grading .....................................................................................................13 2.2 Digital Image Analysis............................................................................................16 2.2.1 Image Representation ...........................................................................................18 2.2.2 Image Processing ..................................................................................................20 2.2.2.1 Image Pre-processing....................................................................................21 2.2.2.2 Image Segmentation .....................................................................................21 2.2.3 Feature Extraction.................................................................................................28 2.2.4 Pattern Classifiers .................................................................................................31 2.2.4.1 Bayesian Classification.................................................................................32 2.2.4.2 Artificial Neural Network.............................................................................33 2.3 Related Works.........................................................................................................38 2.3.1 Image Analysis Application on Agricultural Products.........................................39 ii 3. DESIGN OF ETHIOPIAN COFFEE CLASSIFICATION........................................43 3.1 Overview of Coffee Classification .........................................................................43 3.2 Image Acquisition ...................................................................................................45 3.3 Image Processing.....................................................................................................49 3.4 Feature Extraction..................................................................................................51 3.4.1 Morphological Features ........................................................................................52 3.4.2 Color Features.......................................................................................................53 3.5 Classification Model ...............................................................................................54 3.5.1 Feature Representation .........................................................................................54 3.5.2 Overview of Training and Testing Process...........................................................57 3.5.3 Classifier ...............................................................................................................59 4. IMPLEMENTATION OF COFFEE CLASSIFICATION.........................................62 4.1 Development Environment ....................................................................................62 4.2 Image Binary Analysis............................................................................................63 4.3 Morphology Analysis..............................................................................................66 4.4 Color Analysis .........................................................................................................67 4.5 Experimental Results..............................................................................................70 4.5.1 Naïve Bayes Classifier..........................................................................................71 4.5.2 Neural Network Classifier ....................................................................................78 4.6 Discussion ................................................................................................................87 4.7 Limitations...............................................................................................................89 5. CONCLUSION AND RECOMMENDATIONS..........................................................90 5.1 Conclusion ...............................................................................................................90 5.2 Recommendations...................................................................................................91