請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/63310
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 曹恆偉 | |
dc.contributor.author | Chung-Che Chiu | en |
dc.contributor.author | 邱仲哲 | zh_TW |
dc.date.accessioned | 2021-06-16T16:34:10Z | - |
dc.date.available | 2014-01-16 | |
dc.date.copyright | 2013-01-16 | |
dc.date.issued | 2012 | |
dc.date.submitted | 2012-11-23 | |
dc.identifier.citation | Bibliography
[1] Super-Resolution Imaging. New York: Kluwer Academic Publishers, 2002. [2] Image noise reduction : principles, news, and implications for film making. Available: http://www.nexyad.net/news/NL2-en.html [3] (2010). What is a Sensor? Available: http://sensorcleaning.com/whatisasensor.php [4] S. Baker and T. Kanade, 'Limits on super-resolution and how to break them,' Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol. 24, pp. 1167-1183, 2002. [5] D. Glasner, S. Bagon, and M. Irani, 'Super-resolution from a single image,' in Computer Vision, 2009 IEEE 12th International Conference on, 2009, pp. 349-356. [6] Y. Min-Chun, H. De-An, T. Chih-Yun, and Y. C. F. Wang, 'Self-Learning of Edge-Preserving Single Image Super-Resolution via Contourlet Transform,' in Multimedia and Expo (ICME), 2012 IEEE International Conference on, 2012, pp. 574-579. [7] S. Farsiu, M. D. Robinson, M. Elad, and P. Milanfar, 'Fast and robust multiframe super resolution,' Image Processing, IEEE Transactions on, vol. 13, pp. 1327-1344, 2004. [8] P. Milanfar, Image super-resolution: Historical overview and future challenges: CRC Press, 2010. [9] H. Ozdemir and B. Sankur, 'Assessment of single-frame resolution enhancement algorithms,' in Signal Processing and Communications Applications Conference, 2009. SIU 2009. IEEE 17th, 2009, pp. 145-148. [10] T. Shen-Chuan, K. Tse-Ming, I. Chon-Hong, and L. Tzu-Wen, 'A Fast Algorithm for Single Image Super Resolution in Both Wavelet and Spatial Domain,' in Computer, Consumer and Control (IS3C), 2012 International Symposium on, 2012, pp. 702-705. [11] P. Sung Cheol, P. Min Kyu, and K. Moon Gi, 'Super-resolution image reconstruction: a technical overview,' Signal Processing Magazine, IEEE, vol. 20, pp. 21-36, 2003. [12] R. Fattal, 'Image upsampling via imposed edge statistics,' ACM Trans. Graph., vol. 26, p. 95, 2007. [13] W. T. Freeman, T. R. Jones, and E. C. Pasztor, 'Example-based super-resolution,' Computer Graphics and Applications, IEEE, vol. 22, pp. 56-65, 2002. [14] S. Jian, Z. Jiejie, and M. F. Tappen, 'Context-constrained hallucination for image super-resolution,' in Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on, 2010, pp. 231-238. [15] C.-Y. Yang, J.-B. Huang, and M.-H. Yang, 'Exploiting Self-similarities for Single Frame Super-Resolution,' in Computer Vision – ACCV 2010. vol. 6494, R. Kimmel, R. Klette, and A. Sugimoto, Eds., ed: Springer Berlin Heidelberg, 2011, pp. 497-510. [16] NVIDIA CUDA Compute Unified Device Architecture Programming Guide, 1.1 ed.: NVIDIA Corporation, 2007. [17] R. J. Hovland, 'Latency and Bandwidth Impact on GPU-systems,' Tech report, Norwegian University of Science and Technology, 2008. [18] V. Garcia, E. Debreuve, and M. Barlaud, 'Fast k nearest neighbor search using GPU,' in Computer Vision and Pattern Recognition Workshops, 2008. CVPRW '08. IEEE Computer Society Conference on, 2008, pp. 1-6. [19] S. Arya and D. M. Mount, 'Approximate nearest neighbor queries in fixed dimensions,' presented at the Proceedings of the fourth annual ACM-SIAM Symposium on Discrete algorithms, Austin, Texas, United States, 1993. [20] H. Gunadi, 'Comparing Nearest Neighbor Algorithms in High-Dimensional Space,' College of Engineering and Computer Science, Australian National University, 2011. [21] M. Datar, N. Immorlica, P. Indyk, and V. S. Mirrokni, 'Locality-sensitive hashing scheme based on p-stable distributions,' presented at the Proceedings of the twentieth annual symposium on Computational geometry, Brooklyn, New York, USA, 2004. [22] G. Shakhnarovich, P. Viola, and T. Darrell, 'Fast pose estimation with parameter-sensitive hashing,' in Computer Vision, 2003. Proceedings. Ninth IEEE International Conference on, 2003, pp. 750-757 vol.2. [23] D. Martin, C. Fowlkes, D. Tal, and J. Malik, 'A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics,' in Computer Vision, 2001. ICCV 2001. Proceedings. Eighth IEEE International Conference on, 2001, pp. 416-423 vol.2. [24] W. Zhou, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, 'Image quality assessment: from error visibility to structural similarity,' Image Processing, IEEE Transactions on, vol. 13, pp. 600-612, 2004. [25] A. Barleanu, V. Baitoiu, and A. Stan, 'Floating-point to fixed-point code conversion with variable trade-off between computational complexity and accuracy loss,' in System Theory, Control, and Computing (ICSTCC), 2011 15th International Conference on, 2011, pp. 1-6. | |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/63310 | - |
dc.description.abstract | 摘要
超解析度影像演算法可以把相機記錄下的低解析度影像重建出高解析度的影像,重建的影像會帶有高頻的成分,因此會有更多的細節及更清晰的影像。 超解析度的影像可以應用於兩大領域: ﹝一﹞提升人眼觀測到的圖像品質、﹝二﹞幫助機器自動辨識圖形。 由Glasner提出的「單一影像超解析度演算法」是目前眾多超解析度影像演算法中最有潛力的一種,它不需要額外的資料庫,所計算出的超解析度影像也有不錯的品質,但是這種演算法的運算量很大導致實際運用上的困難。 運算量很大的主要原因是此演算法的核心需要使用到「k位最近鄰居演算法」。 我們針對這個問題設計了「圖像特性分類法」以降低運算量,並且使用平行計算晶片來加速「k位最近鄰居演算法」的計算時間。此系統建立於MATLAB的環境下,並用CUDA C撰寫「k位最近鄰居法」加速的部分。並且我們使用「柏克萊影像分割數據庫」來作為測試比較的基準。 實驗的結果顯示我們所提出的方法最終可以使「單一影像超解析度演算法」達到150倍加速。根據測試結果統計,輸出的超解度影像在PSNR方面只會略微下降0.038db,而結構相似指標(SSIM)則只會下降0.009分。經過實驗證明我們所提出的「圖像特性分類法」配合平行計算晶片加速確實可以加快「單一影像超解析度演算法」,同時可以保證輸出的品質維持不變。我們所提出的改良架構不僅試用於「單一影像超解析度演算法」,另外也可以加速相關的碎形超解析度演算法。所提出的「圖像特性分類法」也有從硬體實現的角度加以設計,輔以定點數模擬來保證系統的一致性。我們提出的系統架構十分地具有潛力,加以發展未來可以使超解析度影像演算法實際應用在真實世界。 | zh_TW |
dc.description.abstract | Abstract
Super resolution imaging is the technology of reconstructing high resolution images with high frequency details from low resolution images recorded by cameras. The needs for high image resolution stem from two application areas: (1) improvement of pictorial information for interpretation; (2) helping representation for automatic machine perception. “Super Resolution from a Single Image” proposed by Glasner is the most promising method among various super resolution approaches, but its computation time is very long due to high dimensional k-nearest-neighbor search. We proposed a novel patch characteristic hashing method with GPU accelerating k-nearest neighbor search to speed-up the process. Our system is implemented on MATLAB, and we use CUDA C to implement KNN search. The proposed architecture is tested with Berkeley Segmentation Dataset and Benchmark. The results show that our method can speed-up “Super Resolution from a Single Image” by 150 times faster. The average PSNR is only 0.038dB lower and Structural Similarity (SSIM) only drops by 0.009. The results implicate that the proposed patch characteristic hashing (PCH) can accelerate “Super Resolution from a Single Image” without affecting output quality of the reconstructed images. | en |
dc.description.provenance | Made available in DSpace on 2021-06-16T16:34:10Z (GMT). No. of bitstreams: 1 ntu-101-R99943133-1.pdf: 2844215 bytes, checksum: f91b87097fa15a5ff9c60463a5d58054 (MD5) Previous issue date: 2012 | en |
dc.description.tableofcontents | Contents
摘要 i Abstract ii Contents iii List of Figures vi List of Tables ix Chapter 1 Introduction 1 1.1 Introduction of Super Resolution 1 1.2 Challenges of Super Resolution 5 1.3 Research Topics and Main Contribution 6 1.4 Thesis Organization 7 Chapter 2 Overview of Super Resolution 9 2.1 Observation Model 10 2.2 Interpolation-based Approaches 11 2.3 Statistical Approaches 12 2.4 Example-based Approaches 15 2.5 Fractal-based Approaches 17 2.6 Summary 23 Chapter 3 Accelerate “Super Resolution from a Single Image” by Using NVIDIA CUDA Based Scalable Computation 25 3.1 Overview of NVIDIA Compute Unified Device Architecture(CUDA) 25 3.1.1 Hardware Architecture 25 3.1.2 Programming and Execution Model 27 3.2 Performance Tuning of CUDA Programs 27 3.2.1 Maximum Number of Threads per Block 27 3.2.2 Parallel Reduction 28 3.2.3 Loop Unrolling 32 3.2.4 Texture Memory 33 3.2.5 Arithmetic Instruction 33 3.3 CPU-GPU Co-working Model 35 3.4 K-Nearest Neighbor Search 38 Chapter 4 Proposed Patch Characteristic Hashing K-Nearest Neighbor Search 45 4.1 Concept of Locality Sensitive Hashing 45 4.2 Proposed Patch Characteristic Hashing (PCH) 46 4.3 Implementation of PCH with Splitting and Combining 48 4.4 Concept of Locality Sensitive Hashing 52 Chapter 5 Experiment Results 55 5.1 Experiment Environment 55 5.2 Computation Time Comparison 58 5.3 PSNR and SSIM Comparison 60 5.4 Fixed-point Analysis 63 5.5 Experiment Results 66 Chapter 6 Conclusion and Future Work 77 6.1 Conclusion 77 6.2 Future Work 77 Bibliography 81 | |
dc.language.iso | en | |
dc.title | 利用圖像特性分類於GPU實現加速單一影像超解析度演算法 | zh_TW |
dc.title | A Novel Algorithm of GPU Acceleration for “Super Resolution from a Single Image” Using Patch Characteristic Hashing | en |
dc.type | Thesis | |
dc.date.schoolyear | 101-1 | |
dc.description.degree | 碩士 | |
dc.contributor.oralexamcommittee | 李揚漢,白宏達,郭景致 | |
dc.subject.keyword | 超解析度,圖形加速卡,K-NN搜尋,圖像特性分類法, | zh_TW |
dc.subject.keyword | Super Resolution,GPU,K-NN search,patch characteristic hashing, | en |
dc.relation.page | 84 | |
dc.rights.note | 有償授權 | |
dc.date.accepted | 2012-11-23 | |
dc.contributor.author-college | 電機資訊學院 | zh_TW |
dc.contributor.author-dept | 電子工程學研究所 | zh_TW |
顯示於系所單位: | 電子工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-101-1.pdf 目前未授權公開取用 | 2.78 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。