臺灣博碩士論文加值系統

本論文針對胃部窄頻影像(Narrow band image, NBI)擷取ROI特徵，先對原始影像中的影像編號及黑色邊框去除掉，將感興趣的大小擷取出來並做圖像前處理，前處理部分主要將影像作灰階、反光點去除、平滑化、值方圖等化等等，前處理後再將影像分成好幾個小正方形區塊(Block)，將每個小區塊用局部三值模式(Local ternary pattern, LTP)取出特徵，並分成病變區塊及正常區塊，一起送入支持向量機做訓練及測試，本方法的最高平均精確率(Precision rate)為91%，且可以在圈選的範圍標示出可能為病變的機率大小，最後整合GLCM和LTP特徵提取演算法於同一個Windows應用程式中，醫生可以藉由此應用程式，幫助醫生圈選出可能的病變範圍。

In this thesis, we purpose a method to detect the early gastric cancer in the narrow band imaging (NBI) images of stomach. We determine the local ternary pattern (LTP) features of the image blocks and then design the classifier based on the support vector machine (SVM) to classify whether the image blocks belong to the normal or abnormal regions. First, we pre-process the original image to make the characteristic more visible. Second, we segment the pre-processed image into several blocks. Third, we compute LTP feature from each cell. Finally, we use an SVM model which is trained by LTP features. The experimental results indicate that the precision rate can be as high as 90% on the 60 testing images.

摘要 i
ABSTRACT ii
誌謝 iii
目錄 iv
表目錄 vi
圖目錄 vii
第一章 緒論 1
1.1 研究動機 1
1.2 研究目的 1
1.3 研究方法 2
1.3.1 影像前處理分析 2
1.3.2 建立訓練資料庫 2
1.3.3 APP 應用程式 2
1.4 論文架構 3
第二章 背景和相關研究 4
2.1 窄頻內視鏡成像 4
2.1.1 胃部內視鏡之微表面結構病徵影像分類 5
2.2 局部二值模式 6
2.3 局部三值模式 7
2.3.1 旋轉不變等價模式 8
2.4 支持向量機 9
第三章 研究方法 12
3.1 系統架構流程圖 12
3.2 影像來源 13
3.3 影像前處理 14
3.4 資料庫分類 16
3.5 支持向量機分類器設計 17
第四章 實驗結果 19
4.1 實驗一：LTP取8點在R=1、R=2及R=3時 20
4.1.1 區塊大小40×40 20
4.1.2 區塊大小50×50 22
4.1.3 區塊大小60×60 23
4.2 實驗二：LTP取12點在R=2；取16點在R=3 25
4.2.1 區塊大小40×40 26
4.2.2 區塊大小50×50 26
4.2.3 區塊大小60×60 27
4.3 實驗三：實驗二加入影像前處理步驟 28
4.3.1 區塊大小40×40 29
4.3.2 區塊大小50×50 30
4.3.3 區塊大小60×60 31
4.4 電腦應用程式APP 32
4.5 實驗結果探討 35
4.5.1 綜合實驗結果 35
第五章 結論與未來展望 42
參考文獻 43

[1]林昆霈, “結合不同LBP特徵之向量量化分類器應用於胃部內視鏡影像早期病變區域之偵測,” 國立雲林科技大學電機工程系碩士班, 2015。
[2]陳懷哲, “結合灰階共生矩陣與向量量化技術做胃部放大內視鏡窄頻影像中早期癌症微血管病徵的偵測,” 國立雲林科技大學電機工程系碩士論文, 2016。
[3]陳家陞, “應用放大內視鏡窄頻影像偵測胃部病徵邊界之研究,” 國立雲林科技大學電機工程系碩士論文, 2017。
[4]楊修銘, “以灰階共生矩陣和支援向量機利用影像區塊重疊特性做早期胃癌病變區域之偵測,” 國立雲林科技大學電機工程系碩士班, 2018。
[5]N. Uedo, F. Mitsuhiro, and G. Kenichi, “Role of narrow band imaging for diagnosis of early-stage esophagogastric cancer: current consensus of experienced endoscopists in Asia–pacific region,” Digestive Endoscopy, vol. 23, pp. 58–71, 2011.
[6]Y. Ezoe, M. Muto, and N. Uedo, “Magnifying narrowband imaging is more accurate than conventional white-light imaging in diagnosis of gastric mucosal cancer,” Gastroenterology, vol. 141, no. 6, pp. 2017-2025, 2011.
[7]T. Ojala, “Multiresolution gray-scale and rotation invariant texture classification with local binary patterns,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 24, no. 7, pp. 971 – 987, 2002.
[8]C. K. Tran, T. F. Lee, L. Chang, and P. J. Chao, “Face description with local binary patterns and local ternary patterns: Improving face recognition performance using similarity feature-based selection and classification algorithm,” International Symposium on Computer, Consumer and Control, Taichung, pp. 520-524, 2014.
[9]R. Peyret, A. Bouridane, F, Khelifi, M. A. Tahir, and S. AI-Maadeed. “Automatic classification of colorectal and prostatic histologic tumor images using multiscale multispectral local binary pattern texture features and stacked generalization.” Neurocomputing 275: 83-93, 2018.
[10]M. A. Abuzneid and A. Mahmood, “Enhanced Human Face Recognition Using LBPH Descriptor, Multi-KNN, and Back-Propagation Neural Network.” IEEE Access, 6, 20641-20651, 2018.
[11]S. Wan, H. C. Lee, X. Huang, T. Xu, T. Xu, X. Zeng, Z. Zhang, Y. Sheikine, J. J. Connolly, J. G. Fujimoto, and C. Zhou. “Integrated local binary pattern texture features for classification of breast tissue imaged by optical coherence microscopy.” Medical image analysis, 38, 104-116, 2017.
[12]W. H. Liao, “Region description using extended local ternary patterns,” International Conference on Pattern Recognition, Istanbul, pp. 1003-1006, 2010
[13]W.H. Liao, C.Y. Liu, and M.C. Lin, “Feature description using center-symmetric extended local ternary patterns,” 2014 IEEE International Symposium on Multimedia, Taichung, pp. 94-97, 2014.
[14]B. Ryu, et al. “Local directional ternary pattern for facial expression recognition.” IEEE Transactions on Image Processing 26(12): 6006-6018, 2017.
[15]X. Tan and W. Triggs. “Enhanced local texture feature sets for face recognition under difficult lighting conditions,” IEEE Transactions on image processing, 1635-1650., 2010.
[16]M. Opitz, M. Diem, S. Fiel, F. Kleber, and R. Sablatnig, “End-to-end text recognition using local ternary patterns, MSER and deep convolutional nets,” 2014 11th IAPR International Workshop on Document Analysis Systems, Tours, pp. 186-190, 2014.
[17]J. Ren, X. Jiang and J. Yuan. “Relaxed local ternary pattern for face recognition,” IEEE international conference on image processing (pp. 3680-3684). IEEE, 2013.
[18]S. R. Dubey, S. K. Singh, and R. K. Singh, “Local bit-plane decoded pattern: a novel feature descriptor for biomedical image retrieval.” IEEE Journal of Biomedical and Health Informatics, 20(4), 1139-1147, 2015.
[19]T. Kanesaka, et al. “Computer-aided diagnosis for identifying and delineating early gastric cancers in magnifying narrow-band imaging,” Gastrointestinal Endoscopy 87(5): 1339-1344, 2018.
[20]C. Cortes and V. Vladimir, “Support-vector networks.” Machine learning, 20(3), 273-297, 1995.
[21]C.-C. Chang and C.-J. Lin, “LIBSVM: A library for support vector machines,” ACM Transactions on Intelligent Systems and Technology, vol. 2, no. 3, pp. 1-27, 2011.