臺灣博碩士論文加值系統

摘要 I
ABSTRACT III
致謝 V
目錄 VI
圖目錄 IX
表目錄 XIII
第一章、 緒論 1
1.1 研究動機 1
1.2 系統架構與流程 2
1.3 論文大綱 3
第二章、 相關習知技術與知識 4
2.1 影像退化模型( Image Degradation Model ) 4
2.2 特徵點偵測 5
2.2.1 SIFT(Scale-invariant feature transform) 5
2.2.2 SURF(Speeded Up Robust Features) 7
2.2.3 ORB (Oriented FAST and rotated BRIEF) 8
2.2.4 KAZE 9
2.2.5 AKAZE(Accelerated-KAZE) 13
2.2.6 特徵點偵測總結 15
2.3 變換矩陣 (Transformation Matrix) 16
2.3.1 仿射轉換 (Affine Transform) 16
2.3.2 透視變換 (Perspective Transformation) 17
2.3.3 單應性矩陣(Homography Matrix) 19
2.3.4 變換矩陣總結 19
2.4 卷積神經網路(Convolutional Neural Network) 20
2.4.1 卷積層(Convolutional Layer) 20
2.4.2 激勵函數(Activation Function) 22
2.4.3 池化層(Pooling Layer) 23
2.4.4 全連接層(Fully Connected Layer) 24
2.5 神經網路特徵傳遞方法 24
2.5.1 殘差網路(Residual Network) 25
2.5.2 密集連接網路(Densely Connected Networks) 27
2.5.3 特徵傳遞網路總結 29
2.6 CNN的注意力機制 29
2.6.1 SENet(Squeeze-and-Excitation Networks) 30
2.6.2 CBAM (Convolutional Block Attention Module) 32
2.6.3 注意力機制總結 34
第三章、 相關方法探討 35
3.1 單張影像運動模糊復原相關文獻 35
3.1.1 直線運動模糊畫面之復原方法 36
3.1.2 Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring 41
3.1.3 Deep Stacked Hierarchical Multi-patch Network for Image Deblurring 45
3.2 視訊影像運動模糊復原相關文獻 49
3.2.1 Deep Video Deblurring for Hand-held Cameras 50
3.2.2 基於深度學習之運動模糊影像之復原方法 52
3.2.3 Triple-Adjacent-Frame Generative Network for Blind Video Motion Deblurring 57
第四章、 本系統與方法 62
4.1 影像暫存模組 64
4.2 相鄰幀對齊模組 65
4.2.1 AKAZE特徵點偵測 65
4.2.2 KNN特徵點匹配 68
4.2.3 單應性矩陣計算 70
4.2.4 透視變換之相鄰幀對齊 72
4.3 視訊去模糊神經網路模組 74
4.3.1 單張影像去模糊網路 76
4.3.2 通道填補與串接 79
4.3.3 相鄰幀融合去模糊網路 81
4.3.4 模糊資料集 91
4.3.5 損失函數 92
4.3.6 網路訓練 94
第五章、 實驗結果 96
5.1 實驗設備與環境 96
5.2 實驗結果與分析 97
5.2.1 主觀結果評估與比較 98
5.2.2 客觀結果評估與比較 111
5.2.3 應用方面比較 112
5.2.4 整體性比較 115
第六章、 結論與未來方向 117
6.1 結論 117
6.2 未來方向 118
參考文獻 119

[1] Kundur D and Hatzinakos D, "Blind Image Deconvolution," IEEE Signal Processing Magazine, pp. 43-64, May 1996.
[2] H. Asai, Y. Oyamada, J. Pilet and H. Saito, "Cepstral analysis based blind deconvolution for motion blur," in 2010 IEEE International Conference on Image Processing, Hong Kong, China, pp.1153-1156, Sept 2010.
[3] D. G. Lowe, "Distinctive Image Features from Scale-Invariant Keypoints," International Journal Comput. Vision, Vol. 60, No. 22, Nov. 2004, pp.91-110.
[4] H. Bay, A. Ess, T. Tuytelaars and L. V. Gool, "Surf: Speeded Up Robust Features," in European Conference on Computer Vision, Berlin, Heidelberg, pp.404-417, 2006.
[5] Ethan Rublee, Vincent Rabaud, Kurt Konolige and Gary Bradski, "ORB: An efficient alternative to SIFT or SURF," in 2011 International Conference on Computer Vision, ICCV, Barcelona, Spain, May 2011, pp.2564-2571.
[6] Pablo F. Alcantarilla, Adrien Bartoli and Andrew J. Davison, "KAZE Features," in European Conference on Computer Vision (ECCV), Fiorenze, Italy, October 2012, pp 214-227.
[7] Pablo Fernández Alcantarilla, Jesús Nuevo, Adrien Bartoli, "Fast Explicit Diffusion for Accelerated Features in Nonlinear Scale Spaces," in British Machine Vision Conference (BMVC), Bristol, UK, September 2013.
[8] P. J. Schneider and D. H. Eberly, Geometric Tools for Computer Graphics., Morgan Kaufmann., 2003.
[9] R. I. Hartley and A. Zisserman, Multiple View Geometry in computer vision, Cambridge University Press, 2003.
[10] Y. Lecun, L. Bottou, Y. Bengio and P. Haffner, "Gradient-based learning applied to document recognition," Proceedings of the IEEE, Vol. 86, No. 6, Nov. 1998, pp.2278-2324.
[11] L. Yann and B. Yoshua, Convolutional networks for images, speech, and time-series, MIT Press, 1995.
[12] V. Nair and G. E. Hinton, "Rectified Linear Units Improve Restricted Boltzmann Machines," in the 27th International Conference on International Conference on Machine Learning, 2010.
[13] Gao Huang, Zhuang Liu, Laurens van der Maaten, Kilian Q. Weinberger, "Densely Connected Convolutional Networks," in The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, Hawaii, USA, July 2017, pp.4700-4708.
[14] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun, "Deep Residual Learning for Image Recognition," in The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, Nevada, USA, June 2016, pp.770-778.
[15] Jie Hu and Li Shen and Gang Sun, "Squeeze-and-Excitation Networks," in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, Utah, USA, June 2018, pp. 7132-7141.
[16] Sanghyun Woo , Jongchan Park , Joon-Young Lee, In So Kweon, "CBAM: Convolutional Block Attention Module," in The European Conference on Computer Vision (ECCV), Munich, Germany, September 2018, pp.3-19.
[17] 陳昭和, 王翔麟, 陳聰毅, “直線運動模糊畫面之復原方法,” 於 2017 第十六屆離島資訊技術與應用研討會(ITAOI-2017), 金門, Taiwan, May 19-21, 2017. (pp. 737-741).
[18] S. Nah, T. H. Kim and K. M. Lee, "Deep Multi-Scale Convolutional Neural Network for Dynamic Scene Deblurring," in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, pp.257-265, July 2017.
[19] Tao, Xin and Gao, Hongyun and Shen, Xiaoyong and Wang, Jue and Jia, Jiaya, "Scale-recurrent Network for Deep Image Deblurring," in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, UT, USA, June 2018.
[20] Zhang, Hongguang and Dai, Yuchao and Li, Hongdong and Koniusz, Piotr, "Deep Stacked Hierarchical Multi-Patch Network for Image Deblurring," in The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, USA, USA, pp.5971-5979, June 2019.
[21] S. Zhou, J. Zhang, W. Zuo, H. Xie, J. Pan and J. S. Ren, "DAVANet: Stereo Deblurring With View Aggregation," in Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, USA, 2019.
[22] S. Su, M. Delbracio, J. Wang, G. Sapiro, W. Heidrich and O. Wang, "Deep Video Deblurring for Hand-held Cameras," in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, pp.237-246, July 2017.
[23] 陳昭和、陳聰毅、林盟富、蕭學良、胡武誌、黃登淵、文成康, “基於深度學習之運動模糊影像之復原方法,” 於 第24屆人工智慧與應用研討會(TAAI2019), 高雄, Taiwan, Nov. 21-23, 2019.
[24] Wenlong Liu, Yuejin Zhao, Ming Liu, Weichao Yi, Liquan Dong, Mei Hui, "Triple-adjacent-frame generative network for blind video motion deblurring," Neurocomputing, vol. 376, pp. 153-165, 2020.
[25] C. Harris and M. Stephens, "A Combined Corner and Edge Detector," in the FourthAlvey Vision Conference, Proc, pp.147-151, 1988.
[26] Fischler, Martin A., and Robert C. Bolles., "Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography.," Communications of the ACM, Vol. 24, No. 6, June 1981, pp.381-395.
[27] Cho, K., van Merrienboer, B., Gulcehre, C., Bougares, F., Schwenk, H., and Bengio, Y., "Learning Phrase Representations using RNN Encoder-Decoder for Statistical," arXiv, 2014.
[28] Wang, Zhou, and Alan C. Bovik, "Mean squared error: Love it or leave it? A new look at signal fidelity measures," IEEE signal processing magazine, pp. 98-117, Vol. 26, No. 1 Jan. 2009.
[29] Zhou Wang, A. C. Bovik, H. R. Sheikh and E. P. Simoncelli, “Image quality assessment: from error visibility to structural similarity,” IEEE Transactions on Image Processing, vol. 13, no. 4, April 2004, pp. 600-612.
[30] D.P. Kingma, and L.J. Ba, "Adam: A Method for Stochastic Optimization," in International Conference on Learning Representations (ICLR 2015), San Diego, CA., May 7-9, 2015, pp 13-13.
[31] S. Cho; J. Wang; S. Lee, "Video Deblurring for Hand-held Cameras Using Patch-based Synthesis," ACM Transactions on Graphics (TOG), Vol. 31, No. 4, July. 2012, pp.1-9.
[32] S. Nah et al., "NTIRE 2019 Challenge on Video Deblurring and Super-Resolution: Dataset and Study," in The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, June, 2019.
[33] J. Redmon and A. Farhadi, "YOLOv3: An Incremental Improvement," arXiv, 2018.