請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/73812完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 劉志文(Chih-Wen Liu) | |
| dc.contributor.author | CHEN XIE | en |
| dc.contributor.author | 謝忱 | zh_TW |
| dc.date.accessioned | 2021-06-17T08:10:50Z | - |
| dc.date.available | 2022-08-19 | |
| dc.date.copyright | 2019-08-19 | |
| dc.date.issued | 2019 | |
| dc.date.submitted | 2019-08-15 | |
| dc.identifier.citation | [1] G. S. Lien, C. W. Liu, J. A. Jiang, C. L. Chuang, and M. T. Teng, ' Magnetic control system targeted for capsule endoscopic operations in the stomach design, fabrication, and in vitro and ex vivo evaluations,' IEEE Trans. Biomed. Eng., vol. 59, no. 7, pp. 2068–2079, July 2012.
[2] X. Zabulis, A. A. Argyros, D. Tsakiris . “Lumen detection for capsule endoscopy,” IEEE/RSJ International Conference on Intelligent Robots and Systems, Nice, France, 2008, pp. 3921-3926. [3] Wang, Dan & Xie, Xiang & Li, Guolin & Yin, Zheng & Wang, Zhihua. (2014). A Lumen Detection-Based Intestinal Direction Vector Acquisition Method for Wireless Endoscopy Systems. IEEE transactions on bio-medical engineering. 10.1109/TBME.2014.2365016. [4] C. S. Bell et al., 'Image partitioning and illumination in image-based pose detection for teleoperated flexible endoscopes', Artif. Intell. Med., vol. 59, no. 3, pp. 185-196, 2013. [5] J. Bernal, J. Sánchez, and F. Vilariño, “Towards automatic polyp detection with a polyp appearance model,” Pattern Recognit., vol. 45, no. 9, pp. 3166–3182, 2012. [6] M. Ganz, X. Yang, and G. Slabaugh, “Automatic segmentation of polyps in colonoscopic narrow-band imaging data,” IEEE Trans. Biomed. Eng., vol. 59, no. 8, pp. 2144–2151, 2012. [7] J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic segmentation,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015, pp. 3431–3440. [8] O. Ronneberger, P. Fischer, and T. Brox, “U-Net : Convolutional Networks for Biomedical,” pp. 234–241, 2015. [9] L. Zhang, S. Dolwani, and X. Ye, “Automated polyp segmentation in colonoscopy frames using fully convolutional neural network and textons,” in Annual Conference on Medical Image Understanding and Analysis, 2017, pp. 707–717. [10] L. Lin, Q. Chen, and S. Yan. Network in network[J]. In: arXiv preprint arXiv,2013 1312.4400 [11] G. L. Liu, M. Yang, and L. D. Bourdev. Compressing deep convolutional networks using vector quantization[J]. CoRR. 2014. vol. abs/1412.6115 [12] N. Iandola, S. Han, M. W. Moskewicz, et al. Squeezenet: Alexnet-level accuracy with 50x fewer parameters and 0.5mb model size[J]. arXiv,2016,1602.07360 [13] C. Kerl, J. Sturm, and D. Cremers, Dense visual SLAM for RGBD cameras[C]. IEEE Conference on Intelligent Robots and Systems, Tokyo, Japan,2013, 2100–2106 [14] Nvidia official website [Online]. Available: https://devblogs.nvidia.com/deep-learning-nutshell-history-training/. [15] Arden Dertat, “Applied Deep Learning” [Online]. Available: https://towardsdatascience.com/applied-deep-learning-part-4-convolutional-neural-networks-584bc134c1e2?gi=68c1e303b507 [16] Stanford CS class CS231n [Online]. Available: http://cs231n.github.io/convolutional-networks/#pool [17] M. A. Nielsen, Neural Networks and Deep Learning. Determination Press, 2015. [18] Y. Bengio, P. Simard, P. Frasconi. Learning long-term dependencies with gradient descent is difficult[J]. IEEE Transactions on Neural Networks, 2002, 5(2):157-166 [19] X. Glorot, Y. Bengio. Understanding the difficulty of training deep feedforward neural networks[J]. Journal of Machine Learning Research, 2010, 9: 249-256 [20] K. He, X. Zhang, S. Ren, et al. Deep Residual Learning for Image Recognition[C]. IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 2016,770-778 [21] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. Deep Residual Learning for Image Recognition. 2015. [22] Ross B. Girshick, Jeff Donahue, Trevor Darrell, and Jitendra Malik. “Rich feature hierarchies for accurate object detection and semantic segmentation”. In: CoRR abs/1311.2524 (2013). [23] JosephRedmon,SantoshDivvala,RossGirshick,andAliFarhadi. “Youonlylookonce:Unified,real-timeobjectdetection”.In:arXiv preprintarXiv:1506.02640(2015). [24] JosephRedmon.Darknet:OpenSourceNeuralNetworksinC.http: //pjreddie.com/darknet/.2013. [25] Bernal, J., Sánchez, F. J., Fernández-Esparrach, G., Gil, D., Rodríguez, C., & Vilariño, F. (2015). WM-DOVA maps for accurate polyp highlighting in colonoscopy: Validation vs. saliency maps from physicians. Computerized Medical Imaging and Graphics, 43, 99-111 . [26] Tajbakhsh, N., Gurudu, S.R., Liang, J.: A classification-enhanced vote accumulation scheme for detecting colonic polyps. In: Yoshida, H., Warfield, S., Vannier, M.W. (eds.) ABD-MICCAI 2013. LNCS, vol. 8198, pp. 53–62. Springer, Heidelberg (2013). doi: 10.1007/978-3-642-41083-3_7 [27] O. Ronneberger, P. Fischer, T. Brox, 'U-net: Convolutional networks for biomedical image segmentation', Proc. Med. Image Comput. Comput.-Assisted Intervention, pp. 234-241, 2015. [28] Long, J., Shelhamer, E., Darrell, T.: Fully convolutional networks for semantic segmentation. In: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3431–3440 (2015) [29] Prerak Mody, “List of Semantic Segmentation Models for Autonomous Vehicles” [Online]. Available: https://blog.playment.io/semantic-segmentation-models-autonomous-vehicles/ [30] MPU-9250 Product Specification;2015 [31] Zhang Z. Camera calibration. In: Medioni G. Kang S, editors. Emerging topics in computer Vision, Prentice Hall Professional Technical Reference; 2004, [chapter 2] | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/73812 | - |
| dc.description.abstract | 隨著人工智慧開始引領潮流，電腦視覺在各行各業中得到直接應用。於此同時，深度捲積神經網絡將電腦視覺的研究推進一個新紀元。
本研究提出了一個通過使用YOLO深度學習模型來讓磁控膠囊内視鏡達到自動導航的方法。YOLO是一個最新的目標偵測深度學習模型。在本論文中我們使用KVASIR的腸道圖片資料集來驗證YOLO模型的腸腔偵測效果。 此外我們使用多任務學習模型將YOLO模型與瘜肉判別的全捲積網絡進行結合。多任務學習模型採用的是Root-Branch架構。Root的部分會共享網絡的前端部分用來提取低層的語義信息，這樣可以減少模型的體積同時也能有效地滿足高計算力的要求。Branch的部分會分別針對不同的學習任務來進行高層語義信息的提取。多任務學習模型能夠以快速和高準確率的表現，在消化道中實時的完成腸腔目標偵測以及瘜肉的自動判別。 | zh_TW |
| dc.description.abstract | As artificial intelligence starts leading the trends in various industries, computer vision can be directly applied to all walks of life. Meanwhile, the application of deep convolutional neural network has pushed computer vision research into a new era.
In this paper, a novel navigation based on deep learning for magnetic field control endoscope is proposed, specifically by using deep learning model “You only look once” (YOLO). YOLO is a state-of-the-art, real-time object detection model. We use KVASIR dataset to evaluate the perfomance of lumen detetection. Furthermore, multi-task learning network is used to integrate YOLO with fully connected neural networks(FCN) which aimed for poly segmentation. The structure of multi-task learning network is Root-Branch. The root part shares former network to extract low level semantic information, which decreases the volume of the model and meet the demand of high computing efficiently. The branch part extracts precise high level semantic information for different tasks. Multi-task learning network is able to detect the lumen and segment the polyp in GI tract with high speed and accuracy performance in real time. | en |
| dc.description.provenance | Made available in DSpace on 2021-06-17T08:10:50Z (GMT). No. of bitstreams: 1 ntu-108-R06921088-1.pdf: 4221651 bytes, checksum: ed08a248db4b25c7ea5203e4c60d57b3 (MD5) Previous issue date: 2019 | en |
| dc.description.tableofcontents | Acknowledgement II
中文摘要 1 Abstract 2 Contents 3 List of Figures 6 List of Tables 11 Chapter 1 Introduction 12 1.1 Background Information 12 1.2 Research Motivation and Purpose 14 1.3 Thesis Overview 16 Chapter 2 Capsule Endoscope Overview and Literature Review 17 2.1 Research Overview of Capsule Endoscope 17 2.2 Survey of Magnetic Field Control System of Capsule Endoscope in Taiwan University 18 2.3 MFN Platform 25 2.3.1 Comparison of 1st and 2nd MFN Platform 26 2.4 Research Overview 28 2.4.1 Lumen Detection 28 2.4.2 Polyp Segmentation 30 2.4.3 Neural Network Compression 31 Chapter 3 Deep Learning 33 3.1 Artificial Neural Networks 34 3.2 Convolutional Layer and Activation Function 35 3.3 Pooling Layer and Fully Connected Layer 37 3.4 Backpropagation Algorithm 38 3.5 Exploding and Vanishing Gradients 41 3.6 Deep Residual Networks 42 Chapter 4 Lumen Detection 45 4.1 YOLO 45 4.1.1 YOLO Algorithm 47 4.1.2 YOLO V3 50 4.2 Experiments and Data Processing 53 4.2.1 Hardware and Frameworks 53 4.2.2 Dataset 53 4.2.3 Data Processing 53 4.2.4 Performance Measuring 54 4.2.5 Result 56 Chapter 5 Polyp Segmentation 60 5.1 Fully Convolutional Neural Networks 60 5.2 Experiments 62 5.2.1 Dataset 62 5.2.2 Performance Measuring 63 5.2.3 Training Details 63 5.2.4 Result 65 Chapter 6 Multi-Task Learning Network 67 6.1 Model Structure 68 6.2 Methodology 70 6.3 Experiments and Results 71 6.3.1 Training Details 71 6.3.2 Results 74 Chapter 7 Lumen Detection Experiment 76 7.1 Experiment Environment 76 7.1.1 Large Intestine Model 76 7.1.2 Capsule Endoscope and Magnetron Large Magnet 77 7.1.3 Magnetic Field Navigator Platform 78 7.2 Camera Calibration 79 7.2.1 Methodology 79 7.2.2 Inertial Measurement Unit 81 7.3 Experiments 83 Chapter 8 Conclusions and Future Works 89 8.1 Conclution 89 8.2 Future Works 90 References 92 | |
| dc.language.iso | en | |
| dc.subject | 膠囊?視鏡 | zh_TW |
| dc.subject | 腸腔偵測 | zh_TW |
| dc.subject | 瘜肉判別 | zh_TW |
| dc.subject | 深度學習 | zh_TW |
| dc.subject | 多任務學習網絡 | zh_TW |
| dc.subject | 電腦視覺 | zh_TW |
| dc.subject | 人工智慧 | zh_TW |
| dc.subject | Multi-task learning network | en |
| dc.subject | Computer vision | en |
| dc.subject | Artificial neural Networks | en |
| dc.subject | Deep learning | en |
| dc.subject | Lumen detection | en |
| dc.subject | Polyp segmentation | en |
| dc.subject | Capsule endoscope | en |
| dc.title | 基於電腦視覺和深度學習的磁控膠囊內視鏡在腸道內自動導航及息肉偵測之研究 | zh_TW |
| dc.title | Study of magnetic controlled capsule endoscope autonomous navigation and polyp segmentation in gastrointestinal tracts based on computer vision and deep learning | en |
| dc.type | Thesis | |
| dc.date.schoolyear | 107-2 | |
| dc.description.degree | 碩士 | |
| dc.contributor.oralexamcommittee | 連吉時(Gi-Shih Lien),粟發滿(Fat Moon Suk) | |
| dc.subject.keyword | 膠囊?視鏡,電腦視覺,人工智慧,深度學習,腸腔偵測,瘜肉判別,多任務學習網絡, | zh_TW |
| dc.subject.keyword | Capsule endoscope,Computer vision,Artificial neural Networks,Deep learning,Lumen detection,Polyp segmentation,Multi-task learning network, | en |
| dc.relation.page | 96 | |
| dc.identifier.doi | 10.6342/NTU201903736 | |
| dc.rights.note | 有償授權 | |
| dc.date.accepted | 2019-08-16 | |
| dc.contributor.author-college | 電機資訊學院 | zh_TW |
| dc.contributor.author-dept | 電機工程學研究所 | zh_TW |
| 顯示於系所單位： | 電機工程學系 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-108-1.pdf 未授權公開取用 | 4.12 MB | Adobe PDF |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。