臺灣博碩士論文加值系統

近年來由於科技的進步與發展，高解析度的裝置在我們的生活中越來越普及。然而傳統的影像處理演算法往往針對低解析度影像所設計，以致於當這些演算法被應用於高解析度影像時會遭遇到運算速度上的瓶頸。為了要減少運算時間，我們可以對原始影像進行取樣，然後對取樣後縮小的影像進行影像處理演算法的運算，得到處理後的結果，最後再將結果放大回原來的解析度，以得到接近在原來高解析度影像進行影像處理的結果。如此一來便可以解決運算速度的問題。然而，經過放大後的處理結果往往會面臨失真的問題。為了要解決這個問題，我們提出了一個新的影像縮放演算法，該演算法可以減少影像縮小後再放大的失真程度，並比過去方法更能達到與原始影像接近的視覺品質。相較於傳統的影像縮放演算法，我們的方法使用了超像素資訊來保存影像中物體的邊緣及紋理等的細節訊號。並基於超像素資訊，我們將所保存的細節訊號應用於影像放大，以重構過去方法常見的失真邊緣及紋理。由實驗結果證明，我們的方法除了在數據結果上比其他的演算法優異外，同時在視覺效果上也明顯優於其他的演算法。

Most traditional image/video processing methods have the computational efficiency bottleneck when they are applied to high-resolution images. To reduce the computation time, the original image is downsampled to a low-resolution image. Then, the processing algorithm is applied to the downsampled image to obtain the processed results. Finally, the image containing the processed results is upsampled to a new image with the same resolution of the original image. Nevertheless, the image quality of the upsampled image with processed results usually contains aliased results. To solve the problem, we propose a new image resizing method to maintain the quality of the processed results after upsampling. Compared to traditional image resizing methods, which generate blurred low-resolution images, our method can preserve both of the edges and the texture of objects in the image by considering superpixel information. As a result, the boundaries of objects in the downsampled low-resolution image are not blurred. Based on the superpixel information, we store the detail signals of edges and textures for upsampling the low-resolution image to the upsampled images with the same resolution of the original image. Experimental results show that the proposed method can achieve better quantitative and qualitative results compared to the state-of-the-art image resizing approaches.

摘要 i
Abstract ii
Index iii
Figure and Table Index iv
1. Introduction 1
2. Related Work 5
3. Superpixel Based Image Resizing 7
3.1. Spatial Quantization 7
3.2. Non-Iterative MAP Pixel Label Assignment 10
3.3. Superpixel based Image Downsampling 13
3.4. Superpixel based Image Upsampling 13
3.5. Boosting Boundary Segmentation in High-Resolution Images 15
4. Experimental Results 17
4.1. Evaluation Metrics 17
4.2. Parameter Selection 19
4.3. Comparisons of Image Resizing 22
4.4. Boosting Boundary Segmentation in High-Resolution Images 28
5. Conclusion 40
6. References 41

[1]A. M. Bruckstein, M. Elad, and R. Kimmel, "Down scaling for better transform compression," IEEE Trans. Image Process., vol. 12, no. 9, pp. 1132-1144, Sep. 2003.
[2]Y. Tsaig, M. Elad, P. Milanfar, and G. H. Golub, "Variable projection for near-optimal filtering in low bit-rate block coders," IEEE Trans. Circuits Syst. Video Technol., vol. 15, no. 1, pp. 154-160, Jan. 2005.
[3]W. Lin and D. Li, "Adaptive downsampling to improve image compression at low bit rates," IEEE Trans. Image Process., vol. 15, no. 9, pp. 2513-2521, Sep. 2006.
[4]H. Schwarz, D. Marpe, and T. Wiegand, "Overview of the scalable video coding extension of the H.264/AVC standard," IEEE Trans. Circuits Syst. Video Technol., vol. 17, no. 9, pp. 1103-1120, Sep. 2007.
[5]X. Wu, X. Zhang, and X. Wang, "Low bit-rate image compression via adaptive down-sampling and constrained least squares upconversion," IEEE Trans. Image Process., vol. 18, no. 3, pp. 552-561, Mar. 2009.
[6]J. Dong and Y. Ye, "Adaptive downsampling for high-definition video coding," IEEE Trans. Circuits Syst. Video Technol., vol. 24, no. 3, pp. 480-488, March 2014.
[7]R. J. Wang, C. W. Huang and P. C. Chang, "Adaptive downsampling video coding with spatially scalable rate-distortion modeling," IEEE Trans. Circuits Syst. Video Technol., vol. 24, no. 11, pp. 1957-1968, Nov. 2014.
[8]G. Farneback, "Two-frame motion estimation based on poly expansion," in Proc. Scandinavian Conference on Image Analysis, 2003.
[9]H. Wang and C. Schmid, "Action recognition with improved trajectories," in Proc. IEEE Conf. Comput. Vis., pp. 3551-3558, 2013.
[10]K. Avgerinakis, K. Adam, A. Briassouli and Y. Kompatsiaris, "Moving camera human activity localization and recognition with motionplanes and multiple homographies," in Proc. IEEE Conf. Image Process., pp. 2085-2089, 2015.
[11]H. Ullah, M. Ullah, H. Afridi, N. Conci and F. G. B. De Natale, "Traffic accident detection through a hydrodynamic lens," in Proc. IEEE Conf. Image Process., pp. 2470-2474, 2015.
[12]A. Stoian, M. Ferecatu, J. Benois-Pineau and M. Crucianu, "Fast action localization in large scale video archives," accepted by IEEE Trans. Circuits Syst. Video Technol., 2015.
[13]Y. G. Jiang, Q. Dai, W. Liu, X. Xue and C. W. Ngo, "Human action recognition in unconstrained videos by explicit motion modeling," IEEE Trans. Image Process., vol. 24, no. 11, pp. 3781-3795, Nov. 2015.
[14]F. Zhou, and F. de la Torre, "Spatio-temporal matching for human pose estimation in video," accepted by IEEE Trans. Pattern Anal. Mach. Intell., 2016.
[15]Y. Zhang, D. Zhao, J. Zhang, R. Xiong, and W. Gao, " Interpolation-dependent image downsampling," IEEE Trans. Image Process., vol. 20, no. 11, pp. 3291-3296, Nov. 2011.
[16]M. Lee and Y. W. Tai, "Robust all-in-focus super-resolution for focal stack photography," IEEE Trans. Image Process., vol. 25, no. 4, pp. 1887-1897, April 2016.
[17]P. Smith, "Bilinear interpolation of digital images," Ultramicroscopy, vol. 6, no. 2, pp. 201-204, 1981.
[18]C. Ren, X. He, Q. Teng, Y. Wu and T. Q. Nguyen, "Single image super-resolution using local geometric duality and non-local similarity," IEEE Trans. Image Process., vol. 25, no. 5, pp. 2168-2183, May 2016.
[19]R. G. Keys, "Cubic convolution interpolation for digital image processing," IEEE Trans. Acoust., Speech, Signal Process., vol. 29, no. 6, pp. 1153-1160, Dec. 1981.
[20]S. K. Park and R. A. Schowengerdt, "Image reconstruction by parametric cubic convolution," Comput. Vis., Graph., Image Process., vol. 23, no. 3, pp. 258-272, Sep. 1983.
[21]J. Shi and S. E. Reichenbach, "Image interpolation by two-dimensional parametric cubic convolution," IEEE Trans. Image Process., vol. 15, no. 7, pp. 1857-1870, Jul. 2006.
[22]S.-H. Hong, L. Wang, T.-K. Truong, T.-C. Lin, and L.-J. Wang, "Novel approaches to the parametric cubic-spline interpolation," IEEE Trans. Image Process., vol. 22, no. 3, pp. 1233-1241, Mar. 2013.
[23]D. Zhou, "An edge-directed bicubic interpolation algorithm," in Proc. 3rd Int. Congr. Image Signal Process., pp. 1186-1189, 2010.
[24]D. Zhou, X. Shen, and W. Dong, "Image zooming using directional cubic convolution interpolation," IET Image Process., vol. 6, no. 6, pp. 627-634, Aug. 2012.
[25]J. K. Chae, J. Y. Lee, M. H. Lee, J. K. Han, T. Q. Nguyen and W. Y. Yeo, "Cubic convolution scaler optimized for local property of image data," IEEE Trans. Image Process., vol. 24, no. 12, pp. 4796-4809, Dec. 2015.
[26]Q. Wang and R. Ward, "A new orientation-adaptive interpolation method," IEEE Trans. Image Process., vol. 16, no. 4, pp. 889-900, Apr. 2007.
[27]L. Zhang and X. Wu, "An edge-guided image interpolation algorithm via directional filtering and data fusion," IEEE Trans. Image Process., vol. 15, no. 8, pp. 2226-2238, Aug. 2006.
[28]X. Li and M. T. Orchard, "New edge-directed interpolation," IEEE Trans. Image Process., vol. 10, no. 10, pp. 1521-1527, Oct. 2001.
[29]X. Zhang, S. Ma, Y. Zhang, L. Zhang, and W. Gao, "Nonlocal edge-directed interpolation," in Proc. IEEE Pacific-Rim Conf. Multimedia, pp. 1197-1207, Dec. 2009.
[30]T. Frajka and K. Zeger, "Downsampling dependent upsampling of images," Signal Process. Image Commun., vol. 19, no. 3, pp. 257-265, Mar. 2004.
[31]J. Goutsias and H. J. A. M. Heijmans, "Nonlinear multiresolution signal decomposition schemes—Part I: Morphological pyramids," IEEE Trans. Image Process., vol. 9, no. 11, pp. 1862-1876, Nov. 2000.
[32]H. J. A. M. Heijmans and J. Goutsias, "Nonlinear multiresolution signal decomposition schemes—Part II: Morphological wavelets," IEEE Trans. Image Process., vol. 9, no. 11, pp. 1897-1913, Nov. 2000.
[33]Univ. Southern California. The USC-SIPI image database. [Online]. Available: http://sipi.usc.edu/database/, accessed Apr. 30, 2016.
[34]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 Proc. 8th IEEE Int. Conf. Comput. Vis., pp. 416-423, Jul. 2001.
[35]H. Nemoto, P. Hanhart, P. Korshunov, and T. Ebrahimi. "Ultra-Eye: UHD and HD images eye tracking dataset." in Proc. IEEE International Workshop on Quality of Multimedia Experience, Sept. 2014.
[36]Y. Wang, P. M. Jodoin, F. Porikli, J. Konrad, Y. Benezeth and P. Ishwar, "CDnet 2014: An Expanded Change Detection Benchmark Dataset," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. Workshops, pp. 393-400, 2014.
[37]Q. Huynh-Thu and M. Ghanbari, "Scope of validity of PSNR in image/video quality assessment," Electron. Lett., vol. 44, no. 13, pp. 800-801, Jun. 2008.
[38]Z. Wang, A. C. Bovik, H. R. Sheikh and E. P. Simoncelli, "Image quality assessment: from error visibility to structural similarity," IEEE Trans. Image Process., vol. 13, no. 4, pp. 600-612, April 2004.
[39]O. Barnich and M.-V. Droogenbroeck, "ViBe: A universal background subtraction algorithm for video sequences," IEEE Trans. Image Process., vol. 20, no. 6, pp. 1709-1724, Jun. 2011.
[40]N. Goyette, P. Jodoin, F. Porikli, J. Konrad, and P. Ishwar, "Changedetection.net: A new change detection benchmark dataset," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. Workshops, pp. 1-8, Jun. 2012.
[41]S. Brutzer, B. Hoferlin, and G. Heidemann, "Evaluation of background subtraction techniques for video surveillance," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., pp. 193-1944, Jun. 2011.