# 臺灣博碩士論文加值系統

(44.200.171.74) 您好！臺灣時間：2022/08/12 07:36

:::

### 詳目顯示

:

• 被引用:0
• 點閱:273
• 評分:
• 下載:0
• 書目收藏:2
 進行紙本手寫字元辨識的研究時，辨識率高低通常與特徵擷取好壞有著密切的關聯。因此在篇本論文當中，將研究著重於特徵的擷取與辨識模組的設計上，希望透過簡單的演算法則，不需經由複雜的數學公式或遞回的處理，就能快速將所需的特徵擷取成功，以提供後續的辨識模組所使用。 本篇論文利用結構式特徵的方法，處理紙本手寫阿拉伯數字與大小寫英文字母之辨識。本文設計的影像辨識系統包含了「影像預前處理」、「特徵擷取」、「字元分類」、「字元辨識」四個階段。其中最重要的是藉由四類特徵，分別為「筆劃端點個數」、「筆劃跨越數」、「筆劃跨越距離」、「筆劃跨越位置」的擷取，計算出每個手寫字元的特徵值。利用這些特徵值的組合，可成功的辨認紙本上手寫的英數字。 本篇論文的研究目標，希望能提出一個簡單且辨識率不錯的辨識演算法，以達到離線手寫字元的辨識目的。透過我們的辨識演算法則，我們請二十位測試者書寫62英數字，包含了10個阿拉伯數字、26大寫英文字母、26個小寫英文字母，總共1240個字，其辨識結果正確率為93.38%，若是除去測試者中書寫草寫小寫字母的因素，那整體的辨識率可提升至97.25%。
 When we research into the off-line handwritten alphanumeric characters, we find that there is a closely relationship between the recognition rate and feature extractions. Therefore, this study is emphasized by the character feature extraction and the recognition models design. It is quick to implement this recognition models because the four features we extracted do not need to be calculated by the complex mathematic formulas and be operated by the difficult recursive algorithm. In this paper, we take the structural features skills to apply to distinguish the off-line handwritten alphanumeric characters including the digits, the uppercase letters and the lowercase letters. This recognition process consists of the following stages, data preprocessing, feature extraction, character classification and character recognition. The most importance of all is to utilize four features, the BPFs, CCFs, CDFs and CPFs, in the classification and recognition procedure to match the characters. By the combination of the mentioned features, we can easily recognize the off-line handwritten alphanumeric characters. The purpose of this study is designed to get reasonable speed and high accuracy to successfully recognize the handwritten alphanumeric characters. Taking this recognition system, we made a simulation on twenty testers, each of them writing 62 characters comprising the digits, the uppercase letters and the lowercase letters. From the results of the simulation, the recognition rate in these 1240 characters is accurately up to 93.38%. However, if we do not consider the script lowercase letters in the test, the recognition rate will rise to 97.25%.
 中文摘要 i ABSTRACT ii ACKNOWLEDGMENTS iii TABLE OF CONTENTS iv LIST OF FIGURES vi LIST OF TABLES vii CHAPTER 1. INTRODUCTION 1 1.1 Motivation 1 1.2 Survey of Related Research 2 1.3 Problem Definition 3 1.4 System Description and Assumptions 3 1.5 Thesis Organization 5 CHAPTER 2. BACKGROUND KNOWLEDGE 7 2.1 History 7 2.2 Regions of Character Recognition 8 2.3 Feature extraction in handwritten Recognition 9 CHAPTER 3. DATA PREPROCESSING 13 3.1 Image Transformation 13 3.2 Character Processing 16 3.3 Character Thinning 19 CHAPTER 4. FEARTUE EXTRACTION 22 4.1 Extraction of Border-points Features 23 4.2 Extraction of Crossing-count Features 24 4.3 Extraction of Crossing-distance Features 26 4.4 Extraction of Crossing-position Features 28 4.5 Character Classification 30 CHAPTER 5. CHARACTER RECOGNITION 32 5.1 Recognition through the BPFs 33 5.2 Recognition through the CCFs 36 5.3 Recognition through the CDFs 37 5.4 Recognition through the CPFs 38 5.5 Ambiguous Cases 39 5.6 Process of the Character Recognition 42 CHAPTER 6. SIMULATION RESULTS 46 6.1 Simulation Results 46 6.2 Simulation Analyses 46 6.3 Summary 48 CHAPTER 7. CONCLUSIONS AND FUTURE WORK 50 7.1 Conclusions 50 7.2 Future works 51 APPENDIX 52 REFERENCES 54
 [1] X. Li, “On-line Handwritten Alphanumeric Character Recognition Using Dominant Points in Strokes”, Master thesis, Department of Computer Science, Hong Kong University of Science and Technology, Hong Kong, 1996.[2] Hinton, G. E., M. Revow, and P. Dayan, “Recognizing Handwritten Digits Using Mixtures of Linear Models”, Advances in Neural Information Processing Systems, Vol. 7, pp.1015-1022, 1995.[3] Dunn, C. E. and P. S. P. Wang, “Character Segmentation Techniques for Handwritten Test — A Survey”, In Proceedings of the 11th International Conference on Pattern Recognition, Vol. 2, pp.577-580, 1992.[4] J. Flusser, and T. Suk, “Pattern Recognition by Affine Moment Invariants”, Pattern Recognition, Vol. 26, pp.167-174, 1993.[5] Liu, K, Y.Q. Cheng, and J, Y. yang, “Algebraic Feature Extraction for Image Recognition Based on Optimal Disriminant Criterion”, Pattern Recognition, Vol 26, No. 6, pp903-911, 1993.[6] K.W. Cheung, “Bidirectional Deformable Matching with Application to Handwritten Character Extraction”, Master thesis, Department of Computer Science, Hong Long Baptist University, Hong Kong, 2000.[7] Cheng, Y. Q., K. Liu, and J.Y. Yang, “A Novel Feature Extraction Method for Image Recognition Based on Similar Discriminant Function”, Pattern Recognition, Vol. 26, pp.115-125, 1993.[8] S.D. Connell, “Writer Adaptation for Online Handwriting Recognition”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 3, pp.329-346, 2002.[9] Z.B. Xu and C.P. Kwong, “Global Convergence and Asymptotic Stability of Asymmetric Hopfield Neural Networks”, Journal of Mathematical Analysis and Applications, Vol. 191, No. 3, pp.405-427, 1995.[10] R.M. Golden, “Exploring the Diversity of Artificial Neural Network Architectures. Review of Neural Networks: A Comprehensive Foundation, by Simon Haykin”, Journal of Mathematical Psychology, Vol. 41, No. 3, pp.287-292, 1997.[11] Y.J. Chang “Texture Classification Using C-matrix and Fuzzy Min-max Neural Network”, Master thesis, Department of Information and Computer Engineering, Chung Yuan Christian University, Taiwan, ROC, 1996.[12] C.M. Travieso and C.R Morales, “Handwritten Digits Parameterisation for HMM Based Recognition”, Image Processing and Its Application, Conference Publication, No.465, pp.770-774, 1999.[13] K.F. Chan and D.Y. Yeung, “Recognizing On-line Handwritten Alphanumeric Characters through Flexible Structural Matching”, Pattern Recognition, Vol. 32, pp.1099-1114, 1998.[14] Y. Lu, “Machine Printed Character Segmentation — An Overview”, Pattern Recognition, Vol. 28, No. 1, pp.67-80, 1995.[15] N. Arica and T.Y. Vural, “One Dimensional Representation of Two Dimensional Information for HMM Based Handwritten Recognition”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 20, No. 2, pp.948-952, 1998.[16] D.E. Goldberg, “Genetic Algorithms in Search”, Optimization and Machine Learnig Reading, Adddison-Wesley publishing company, 1989.[17] R.L. Haupt and S.E. Haupt, Genetic Algorithms, Springer publishing company, 1999.[18] LeCun, Y., B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard, and L.D. Jackel, “Backpropagation Applied to Handwritten Zip Code Recognition”, Neural Computation, Vol. 1, pp.541-551, 1989.[19] Suen, C. Y. R. Legault, C. Nadal, M. Cheriet, and L. Lam, “Building a New Generation of Handwriting Recognition Systems”, Pattern Recognition Letters, Vol. 14, No. 4, pp.303-315, 1993.[20] Gosselin, (1993), MULTITEL-TCTS LABhttp://tcts.fpms.ac.be/rdf/hcrgenuk.htm[21] Herman, G.. T. and H. K. Liu, “Dynamic Boundary Surface Detection”, Computer Graphics and Image Processing, Vol. 7, No. 1, pp.130-138, 1978.[22] Lam, L. and C. Y. Suen, “Optimal Combinations of Pattern Classifiers”, Pattern Recognition Letters, Vol. 16, pp. 945-954, 1995.[23] R.C. Gonzalez and R.E. Woods, Digit Image Processing, Adddison-Wesley publishing company, 1993.
 國圖紙本論文
 推文當script無法執行時可按︰推文 網路書籤當script無法執行時可按︰網路書籤 推薦當script無法執行時可按︰推薦 評分當script無法執行時可按︰評分 引用網址當script無法執行時可按︰引用網址 轉寄當script無法執行時可按︰轉寄

 無相關論文

 1 4. 白蓉生, 1996, ”增層法多層板與非機械鑽孔式導孔”, 工業材料118期, pp.98.

 1 手寫紙本是非及選擇題答案辨識系統之實作 2 紙本手寫國語注音符號辨識系統之研究 3 網路系統之模糊語意問卷設計與分析 4 網際網路即時測驗平台上之資料區塊排程策略於延遲公平性與頻寬最小化之研究 5 倒傳遞類神經網路在測驗能力值估算之模擬研究 6 電腦化模糊認知診斷系統之研究 7 潛藏馬可夫模型應用於轉換機率之估算正確性模擬研究 8 多重相關加權核平滑化多元計分試題選項關聯結構之研究 9 多元計分三參數試題選項分析模式半最大概似估計法之研究 10 無參數李克特氏模糊計分模式之研究 11 兒童音樂創造思考測驗編製研究 12 線上題庫與評量系統之開發 13 國民小學學童口語理解量表相關之研究 14 國小學生乘除法表徵能力與後設認知相關之研究 15 國小三年級數學科正整數乘法概念之探究--以試題選項特徵曲線為分析基礎

 簡易查詢 | 進階查詢 | 熱門排行 | 我的研究室