(3.235.108.188) 您好!臺灣時間:2021/03/03 20:58
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:王聖淵
研究生(外文):SHENG-YUAN WANG
論文名稱:基於卷積神經網路之影像去糢糊方法
論文名稱(外文):Convolutional Neural Network for Image Deblurring
指導教授:王家慶
學位類別:碩士
校院名稱:國立中央大學
系所名稱:資訊工程學系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2018
畢業學年度:106
語文別:中文
論文頁數:68
中文關鍵詞:深度學習影像處理影像去糢糊機器學習影像解糢糊
相關次數:
  • 被引用被引用:0
  • 點閱點閱:74
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
近年隨著深度學習的崛起,在學術界、業界每隔幾個月就會有驚人的深度學習成果與作品誕生,同時也證明深度學習技術應用在影像方面有許多不凡的效果。本論文以使用卷積神經網路為主要方法,目的是使因自然光學失焦、拍照時手震等等因素產生的糢糊化影像恢復成清晰影像。本論文提出了三種網路架構:Auto_deblur、S-Net和AGDNet;在圖片輕微受損糢糊的情況下以選擇S-Net為佳,因為S-Net執行速度很快;在對糢糊程度較為嚴重的目標時主要以執行AGDNet效果最好,它整合了前兩種網路的構思與優點;除此之外本論文還提出了在訓練網路時改良損失函數使網路輸出更擬合真實清晰的圖像。本架構除了在解決糢糊化問題上有好的表現外,在超解析度成像(Super-Resolution)、解決圖片雜訊(Image Denoising)和影像恢復(Image Restoration)問題上同時也有好的效果。在實驗過程中,結果也顯示本方法較其他深度類神經網路和業界常用解法表現更加優秀。
In recent years, along with the rise of deep learning in academia and industry. There will be striking deep learning achievements and works every few months. It also proves that deep learning technology application has many great effects in the image. In this paper, the convolution neural network is used as the main method to restore out of focus images or blurred images to clear images. This paper proposes three network architectures: Auto_deblur, S-Net and AGDNet. In the case that the image is slightly damaged and blurred, it is better to select S-Net, because S-Net can execute quickly. AGDNet has the best effect when the image has a relatively serious target, which integrates the conception and advantages of the first two networks. In addition, this paper also proposes to the improved loss function in training the network so that the network output is able to fit more the real and clear images. In addition to its good performance in solving deblurring, this architecture also has good effects in image super-resolution, Image denoising and Image Restoration. The results also show that this method performs better than other deep neural networks and other commonly used solutions in the industry.
中文摘要 I
ABSTRACT II
圖目錄 III
表目錄 V
章節目次 VI
第一章 緒論 1
1.1 研究背景 1
1.2 研究動機與目的 2
1.3 研究方法與章節概要 3
第二章 相關研究 4
第三章 深度學習 10
3.1 類神經網路 11
3.1.1 類神經網路的發展 11
3.1.2 感知機 12
3.1.3 多層感知機 15
3.1.4 倒傳遞演算法 16
3.2 深度學習 23
3.2.1 卷積神經網路(Convolutional Neural Network, CNN) 23
3.2.2 梯度消失和梯度爆炸 25
第四章 去糢糊系統架構 27
4.1 提出架構 27
4.1.1 基於自編碼器(Autoencoder)的去糢糊方法 27
4.1.2 基於SRCNN的去糢糊方法 29
4.1.3 基於我們的去糢糊方法 30
4.2 損失函數(LOSS FUNCTION) 33
4.2.1 像素級損失(Pixel Loss) 36
4.2.2 多層結構相似性損失(MS-SSIM Loss) 36
第五章 實驗設計與實驗結果 38
5.1 電腦軟硬體配置 38
5.2 資料集說明 39
5.3 實驗設計 40
5.3.1 訓練參數 40
5.3.2 訓練模型與網路架構比較 41
5.3.3 實驗度量方式 41
5.4 實驗結果與數據比較 41
5.4.1 提出的三種網路實驗 41
5.4.2 改良AGDNet訓練時的損失函數 44
5.4.3 與估計核(Estimate Kernel)方法的比較實驗 44
5.4.4 與其它類神經網路比較 45
5.5 延伸應用 47
5.5.1 超解析度成像(Super-Resolution) 47
5.5.2 去雜訊(Denoising) 49
第六章 結論與未來研究方向 50
參考文獻 51
[1] Hinton, Geoffrey E., Simon Osindero, and Yee-Whye Teh. "A fast learning algorithm for deep belief nets." Neural computation 18.7 (2006): 1527-1554.
[2] LeCun, Yann, et al. "Gradient-based learning applied to document recognition." Proceedings of the IEEE 86.11 (1998): 2278-2324.
[3] Mikolov, Tomáš, et al. "Recurrent neural network based language model." Eleventh Annual Conference of the International Speech Communication Association. 2010.
[4] Krizhevsky, Alex, Ilya Sutskever, and Geoffrey E. Hinton. "Imagenet classification with deep convolutional neural networks." Advances in neural information processing systems. 2012.
[5] Silver, David, et al. "Mastering the game of go without human knowledge." Nature 550.7676 (2017): 354.
[6] Gatys, Leon A., Alexander S. Ecker, and Matthias Bethge. "A neural algorithm of artistic style." arXiv preprint arXiv:1508.06576 (2015).
[7] Nalbach, Oliver, et al. "Deep Shading: Convolutional Neural Networks for Screen Space Shading." Computer Graphics Forum. Vol. 36. No. 4. 2017.
[8] Adobe MAX 2011 - Photoshop Image Deblurring sneak [Online]. Available: https://www.youtube.com/watch?v=xxjiQoTp864. [Accessed: 22-Jun 2018]
[9] Levin, Anat, et al. "Understanding and evaluating blind deconvolution algorithms." Computer Vision and Pattern Recognition, 2009. CVPR 2009. IEEE Conference on. IEEE, 2009.
[10] Shan, Qi, Jiaya Jia, and Aseem Agarwala. "High-quality motion deblurring from a single image." Acm transactions on graphics (tog). Vol. 27. No. 3. ACM, 2008.
[11] Fergus, Rob, et al. "Removing camera shake from a single photograph." ACM transactions on graphics (TOG). Vol. 25. No. 3. ACM, 2006.
[12] Pan, Jinshan, et al. "Blind image deblurring using dark channel prior." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016.
[13] Cortes, Corinna, and Vladimir Vapnik. "Support-vector networks." Machine learning 20.3 (1995): 273-297.
[14] Quinlan, J. Ross. "Induction of decision trees." Machine learning 1.1 (1986): 81-106.
[15] Ho, Tin Kam. "Random decision forests." Document analysis and recognition, 1995., proceedings of the third international conference on. Vol. 1. IEEE, 1995.
[16] Schalkoff, Robert J. Artificial neural networks. Vol. 1. New York: McGraw-Hill, 1997.
[17] McCulloch, Warren S., and Walter Pitts. "A logical calculus of the ideas immanent in nervous activity." The bulletin of mathematical biophysics 5.4 (1943): 115-133.
[18] Hebb, Donald O. "The organization of behavior: A neuropsychological theory." (1949).
[19] Rosenblatt, Frank. "The perceptron: a probabilistic model for information storage and organization in the brain." Psychological review 65.6 (1958): 386.
[20] Marvin, Minsky, and Papert Seymour. "Perceptrons." (1969).
[21] Rumelhart, David E., Geoffrey E. Hinton, and Ronald J. Williams. "Learning representations by back-propagating errors." nature 323.6088 (1986): 533.
[22] Vanishing gradient problem [Online]. Available: https://en.wikipedia.org/wiki/Vanishing_gradient_problem . [Accessed: 22-Jun 2018]
[23] ImageNet [Online]. Available: https://en.wikipedia.org/wiki/ImageNet . [Accessed: 22-Jun 2018]
[24] 蘇木春、張孝德 編著, 機器學習:類神經網路、糢糊系統以及基因演算法則,第二版全華科技圖書, 民國一百零一年.
[25] Neurotransmitter receptor [Online]. Available: https://en.wikipedia.org/wiki/Neurotransmitter_receptor .[Accessed: 23-Jun 2018]
[26] Haykin, Simon, and Bernard Widrow, eds. Least-mean-square adaptive filters. Vol. 31. John Wiley & Sons, 2003.
[27] Introduction to Different Activation Functions for Deep Learning [Online]. Available: https://medium.com/@shrutijadon10104776/survey-on-activation-functions-for-deep-learning-9689331ba092 . [Accessed: 23-Jun 2018]
[28] O'Shea, Keiron, and Ryan Nash. "An introduction to convolutional neural networks." arXiv preprint arXiv:1511.08458 (2015).
[29] Vincent, Pascal, et al. "Stacked denoising autoencoders: Learning useful representations in a deep network with a local denoising criterion." Journal of Machine Learning Research 11.Dec (2010): 3371-3408.
[30] Ronneberger, Olaf, Philipp Fischer, and Thomas Brox. "U-net: Convolutional networks for biomedical image segmentation." International Conference on Medical image computing and computer-assisted intervention. Springer, Cham, 2015.
[31] He, Kaiming, et al. "Deep residual learning for image recognition." Proceedings of the IEEE conference on computer vision and pattern recognition. 2016.
[32] Mackiewicz, Andrzej, and Waldemar Ratajczak. "Principal components analysis (PCA)." Computers and Geosciences 19 (1993): 303-342.
[33] Goodfellow, Ian, et al. "Generative adversarial nets." Advances in neural information processing systems. 2014.
[34] Dong, Chao, et al. "Learning a deep convolutional network for image super-resolution." European Conference on Computer Vision. Springer, Cham, 2014.
[35] Li, Zuoxin, and Fuqiang Zhou. "FSSD: Feature Fusion Single Shot Multibox Detector." arXiv preprint arXiv:1712.00960 (2017).
[36] What exactly is the degradation problem that Deep Residual Networks try to alleviate? [Online]. Available: https://www.quora.com/What-exactly-is-the-degradation-problem-that-Deep-Residual-Networks-try-to-alleviate . [Accessed: 23-Jun 2018]
[37] Springenberg, Jost Tobias, et al. "Striving for simplicity: The all convolutional net." arXiv preprint arXiv:1412.6806 (2014).
[38] Peak signal-to-noise ratio [Online]. Available: https://en.wikipedia.org/wiki/Peak_signal-to-noise_ratio . [Accessed: 23-Jun 2018]
[39] Wang, Zhou, et al. "Image quality assessment: from error visibility to structural similarity." IEEE transactions on image processing 13.4 (2004): 600-612.
[40] Human visual system model [Online]. Available: https://en.wikipedia.org/wiki/Human_visual_system_model . [Accessed: 23-Jun 2018]
[41] Wang, Zhou, Eero P. Simoncelli, and Alan C. Bovik. "Multiscale structural similarity for image quality assessment." Signals, Systems and Computers, 2004. Conference Record of the Thirty-Seventh Asilomar Conference on. Vol. 2. Ieee, 2003.
[42] Zhao, Hang, et al. "Loss functions for image restoration with neural networks." IEEE Transactions on Computational Imaging 3.1 (2017): 47-57.
[43] Bevilacqua, Marco, et al. "Low-complexity single-image super-resolution based on nonnegative neighbor embedding." (2012): 135-1.
[44] Zeyde, Roman, Michael Elad, and Matan Protter. "On single image scale-up using sparse-representations." International conference on curves and surfaces. Springer, Berlin, Heidelberg, 2010.
[45] Huang, Jia-Bin, Abhishek Singh, and Narendra Ahuja. "Single image super-resolution from transformed self-exemplars." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2015.
[46] Martin, David, et al. "A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics." Computer Vision, 2001. ICCV 2001. Proceedings. Eighth IEEE International Conference on. Vol. 2. IEEE, 2001.
[47] Kingma, Diederik P., and Jimmy Ba. "Adam: A method for stochastic optimization." arXiv preprint arXiv:1412.6980 (2014).
[48] Zhang, Kai, et al. "Learning deep CNN denoiser prior for image restoration." arXiv preprint (2017).
[49] Xu, Li, et al. "Deep convolutional neural network for image deconvolution." Advances in Neural Information Processing Systems. 2014.
[50] scikit-image [Online]. Available: http://scikit-image.org/ . [Accessed: 23-Jun 2018]
[51] Whyte, Oliver, et al. "Non-uniform deblurring for shaken images." International journal of computer vision 98.2 (2012): 168-186.
[52] Pan, Jinshan, et al. "Blind image deblurring using dark channel prior." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016.
[53] Cho, Sunghyun, Jue Wang, and Seungyong Lee. "Handling outliers in non-blind image deconvolution." Computer Vision (ICCV), 2011 IEEE International Conference on. IEEE, 2011.
[54] Krishnan, Dilip, and Rob Fergus. "Fast image deconvolution using hyper-Laplacian priors." Advances in Neural Information Processing Systems. 2009.
[55] Levin, Anat, et al. "Image and depth from a conventional camera with a coded aperture." ACM transactions on graphics (TOG) 26.3 (2007): 70.
[56] Schuler, Christian J., et al. "A machine learning approach for non-blind image deconvolution." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2013.
[57] Why multiply the error by the derivative of the sigmoid in neural networks? [Online]. Available:https://stackoverflow.com/questions/45787261/why-multiply-the-error-by-the-derivative-of-the-sigmoid-in-neural-networks . [Accessed: 28-Jun 2018]
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔