臺灣博碩士論文加值系統

本論文使用可程式化邏輯閘陣列(Field-Programmable Gate Array， FPGA)來實
現雙邊濾波器以進行影像濾波。雙邊濾波器的優勢在於不僅能顯著濾除雜訊，同時
能夠保留影像邊緣資訊，因此被廣泛運用在不同影像處理中，但其計算也較為複雜
而需要大量資源。本論文提出了一個雙邊濾波器的硬體實現方法，經由降低演算法
的複雜度，使雙邊濾波器能達到即時運算。本方法為暫存器級別之平行管線化架構，
使用暫存器陣列達成雙邊濾波器內像素之分組，計算相似度函數後傳入距離分量
及加法樹來加速接近度函數的運算，即時獲得結果，該方法並可輕易改變濾波器核
心大小以達不同濾波效果。該方法實現於Xilinx Zynq 上，其FPGA時脈為150MHz，
實驗與已有方法進行峰值訊噪比和結構相似性指標，證明本方法可以得到較佳濾
波效果。在執行速度上，對於1920x1080 的8 位元灰階影像，本方法比未使用查找
表且運行在個人電腦上的原始濾波器快了574 倍，並可達到33FPS 即時效能。

This thesis applies the FPGA (Field-Programmable Gate Array) to implement real-time bilateral filter for image denoising. The advantages of bilateral filter are to reduce noise effectively but at same time preserve the structural information of images. Therefore, the bilateral filter is widely used in image processing. However, the bilateral filter needs more computational resources because of its complex calculation involved. We propose a method to implement a real-time bilateral filter through reducing computational complexity. Our design is designed on register-level parallelized pipeline architecture, and separating bilateral filter’s input data into groups. After calculating the similarity function, results were transferred to a distance component and then an adder tree. This process is introduced to accelerate the operation of the similarity function and obtain result in real-time. The filter size of this method can be easily adjustable different filtering effects. This method was implemented on Xilinx Zynq, and the frequency of FPGA is 150MHz. We used PSNR (Peak Signal to Noise Ratio) and SSIM (Structural Similarity Index) to validate our design and the comparison with the existing methods with the different filter sizes demonstrates that our design can achieve better image quality. In terms of execution speed, our design is 574 times faster than the classical bilateral filter running on a PC for grey images that are 1920x1080 and have 8-bit depth. Our method can achieve real-time performance with 33 FPS.

摘要 i
英文摘要 ii
誌謝 iii
目錄 iv
表目錄 vi
圖目錄 vii
第1章 前言 1
1.1 研究背景 1
1.2 研究動機 2
1.3 研究目的及方法 2
1.4 論文架構 4
第2章 文獻探討 5
2.1 雙邊濾波器之原理 5
2.2 雙邊濾波器之相關研究 8
2.3 實現於Zynq系統上之影像處理研究 8
2.4 於嵌入式系統上實現雙邊濾波器之相關研究 10
2.4.1 接近度函數 11
2.4.2 相似度函數 12
2.4.3 接近度函數與相似度函數之比較 15
2.5 代表性即時雙邊濾波器之說明 16
第3章 影像處理演算法實現於Zynq系統 23
3.1 Zynq-7000 All Programmable SoC 23
3.2 Zynq SoC開發流程 25
3.3 Zynq工具 28
3.4 Zynq系統用於影像處理之架構與其限制 30
第4章 本論文方法與系統架構 33
4.1 暫存器陣列 33
4.2 相似度函數 36
4.2.1 多項式內插逼近法 37
4.2.2 泰勒展開式逼近法 39
4.2.3 查找表實現泰勒展開式逼近法 43
4.3 接近度函數 44
4.4 系統彈性架構 48
第5章 實驗分析 53
5.1 影像濾波結果比較 54
5.2 運算時間 71
5.3 於Zynq系統上實現之資源消耗 73
第6章 結論 77
參考文獻 79

[1]C. Tomasi and R. Manduchi, "Bilateral Filtering for Gray and Color Images," in Proc. 6th Int. Conf. on Computer Vision, Bombay, pp. 839-846, Jan. 1998.
[2]Durand and J. Dorsey, "Fast Bilateral Filtering for the Display of High-Dynamic-Range Images," ACM Transactions on Graphics, vol. 21, no. 3, pp. 257-266, July 2002.
[3]J. Xie, "Color Image Diffusion Using Adaptive Bilateral Filter," in Proc. 27th Annual Conf. IEEE Engineering in Medicine and Biology, Shanghai, pp. 3433-3436, Jan. 2006.
[4]M. Olfa and K. Nawres, "Ultrasound Image Denoising Using a Combination of Bilateral Filtering and Stationary Wavelet Transform," in Proc. Conf. on Image Processing, Applications and Systems Conference (IPAS), pp. 1-5, Nov. 2014.
[5]W. Zhongce, "A Bilateral Filtering Based Image Denoising Algorithm for Nighttime Infrared Monitoring Images," in Proc. Conf. Computational Science and Computational Intelligence (CSCI), pp. 199-203, Mar. 2014.
[6]H. Lu, Y. Li and S. Serikawa, "Underwater Image Enhancement Using Guided Trigonometric Bilateral Filter And Fast Automatic Color Correction," in Proc. 20th IEEE Int. Conf. on Image Processing (ICIP), pp. 3412-3416, Sep. 2013.
[7]J. Rydell, "Bilateral Filtering of fMRI Data," IEEE Journal of Selected Topics in Signal Processing, vol. 2, no. 6, pp. 891-896, Dec. 2008.
[8]W. G. Zhang, "SAR Image Despeckling via Bilateral Filtering," Electronics Letters, vol. 45, no. 15, pp. 781-783, July 2009.
[9]H. Tabkhi, M. Sabbagh and G. Schirner, "Power-efficient Real-time Solution for Adaptive Vision Algorithms," IET Computers & Digital Techniques, vol. 9, no. 1, pp. 16-26, June 2015.
[10]M. Russell and S. Fischaber, "OpenCV Based Road Sign Recognition on Zynq," in Proc. 11th IEEE Int. Conf. on Industrial Informatics (INDIN), Bochum, pp. 596-601, July 2013.
[11]A. T. Hoang, T. Koide and M. Yamamoto, "Low Cost Hardware Implementation for Traffic Sign Detection System," in Proc. IEEE Asia Pacific Conf. on Circuits and Systems (APCCAS), Ishigaki, pp. 363-366, Nov. 2014.
[12]T. Han, G. Wen Liu, H. Cai and B. Wang, "The Face Detection and Location System Based on Zynq," in Proc. 11th Int. Conf. on Fuzzy Systems and Knowledge Discovery (FSKD), Xiamen, pp. 835-839, Aug. 2014.
[13]V. Suse and D. Ionescu, "A Real-time Reconfigurable Architecture for Face Detection," in Proc. Int. Conf. on ReConFigurable Computing and FPGAs (ReConFig), Mexico City, pp.1-6, Dec. 2015.
[14]F. Schwiegelshohn and M. Hubner, "Design of an Attention Detection System on the Zynq-7000 SoC," in Proc. Int. Conf. on ReConFigurable Computing and FPGAs (ReConFig), Cancun, pp. 1-6, Dec. 2014.
[15]S. Xie, Y. Li, Z. Jia and L. Ju, "Binarization Based Implementation for Real-Time Human Detection," in Proc. Int. Conference on Field Programmable Logic and Applications, pp.13-16, Sep. 2013.
[16]K. Negi, K. Dohi, Y. Shibata and K. Oguri, "Deep Pipelined One-Chip FPGA Implementation of A Real-Time Image-Based Human Detection Algorithm," in Proc. Int. Conference on Field-Programmable Technology, pp. 1-8, Dec. 2011.
[17]M. Hahnle, F. Saxen, M. Hisung, U. Brunsmann and K. Doll, "FPGA-Based Real-Time Pedestrian Detection on High-Resolution Images," in Proc. IEEE Conference on Computer Vision and Pattern Recognition Workshops, pp. 629-635, June 2013.
[18]H. Nguyen. T. K, C. Belleudy and T. V. Pham, "Power Evaluation of Sobel Filter on Xilinx Platform," in Proc. Int. Conf. on Faible Tension Faible Consommation (FTFC), Monaco, pp. 1-5, May 2014.
[19]A. Dallai and S. Ricci, "Real-time Bilateral Filtering of Ultrasound Images through Highly Optimized DSP Implementation," in Proc. Int. Conf. on Education and Research Conference (EDERC), pp. 278- 281, Oct. 2014.
[20]V. Jakhetiya, W. Lin and S. P. Jaiswal, "Observation Model Based Perceptually Motivated Bilateral Filter for Image Reconstruction," in Proc. Int. Conf. on Digital Signal Processing (DSP), pp. 201- 205, Sep. 2015.
[21]V. Patanavijit, " The Bilateral Denoising Performance Influence of Window, Spatial and Radiometric Variance," in Proc. Int. Conf. on Advanced Informatics: Concepts, Theory and Applications (ICAICTA), Bang Saen Beach, pp. 15-20, Nov. 2015.
[22]O. P. Gangwal, E. Coezijn and R. P. Berretty, "Real-time Implementation of Depth Map Post-processing for 3D-TV on A Programmable DSP (TriMedia) ," in Proc. Int. Conf. on Consumer Electronics( ICCE ), pp. 9-10, Jan. 2009.
[23]A.S. Babalis, A.N. Venetsanopoulos and D. Androutsos, " Investigation of the Effect of Three-dimensional Smoothing on Multiview Stereo Images," in Proc. Int. Conf. on Digital Signal Processing (DSP), pp.1-7, July 2011.
[24]H. Dutta, F. Hannig and J. Teich, "A Design Methodology for Hardware Acceleration of Adaptive Filter Algorithms in Image Processing," in Proc. Int. Conf. on Application-specific Systems, Architectures and Processors(ASAP), pp.15-23, Sep. 2006.
[25]C. Charoensak and F. Sattar, "FPGA Design of a Real-time Implementation of Dynamic Range Compression for Improving Television Picture," in Proc. Int. Conf. on Information, Communications & Signal Processing(ICSP), pp. 1-5, May 2007.
[26]W. Luping and Mafeng, "High-Dynamic-Range Infrared Image Enhancement and Hardware Design," in Proc. Int. Conf. on Computer Vision in Remote Sensing (CVRS), pp. 127-131, Dec. 2012.
[27]J. S. S. Kutty, F. Boussaid and A. Amira, "A High Speed Configurable FPGA Architecture for Bilateral Filtering," in Proc. Int. Conf. on Image Processing (ICIP), pp.1248-1252, Oct. 2014.
[28]A. Gabiger-Rose, M. Kube, R. Weigel and R. Rose, "An FPGA-Based Fully Synchronized Design of a Bilateral Filter for Real-Time Image Denoising," IEEE Transactions on Industrial Electronics, vol. 61, no. 8, pp. 4093-4104, Nov. 2013.
[29]T. Q. Vinh , J. H. Park, Y. Kim and S. H. Hong, "FPGA Implementation of Real-Time Edge-Preserving Filter for Video Noise Reduction," in Proc. Int. Conf. on Computer and Electrical Engineering(ICCEE), pp. 611-614, Dec. 2008.
[30]T. W. Roberts, "Non-oscillatory Interpolation for the Semi-Lagrangian Scheme," Computer Physics Communications, vol. 183, no. 5, pp. 1094-1100, May 2012
[31]K. Imajo, "Fast Gaussian Filtering Algorithm Using Splines," in Proc. Int. Conf. on Pattern Recognition (ICPR), pp. 489-492, Nov. 2012.
[32]N. N. Schraudolph, "A Fast, Compact Approximation of the Exponential Function," Neural Computation, vol. 11, no. 4, pp. 853-862, May 1999.
[33]G. Muntingh, "Topics in Polynomial Interpolation Theory," Mathematics of Computation, vol. 52, pp.703-723, Sep. 2011.
[34]T. Yamaguchi, M. Ikehara and Y. Nakajima, "Image Interpolation Based on Weighting Function of Gaussian," in Proc. Int. Conf. on Systems and Computers, Signals, pp. 2481-2499, Nov. 2015
[35]P. Nilsson, A. U. R. Shaik and R. Gangarajaiah, "Hardware Implementation of the Exponential Function Using Taylor Series," in Proc. Int. Conf. on NORCHIP, pp. 1-4, Oct. 2014.
[36]D. Sonowal and M. Bhuyan, " Linearizing Thermistor Characteristics by Piecewise Linear Interpolation in Real Time FPGA," in Proc. Int. Conf. on Advances in Computing, Communications and Informatics, pp. 1976-1980, Aug. 2013.
[37]D. Lee, R. Cheung and W. Luk, "Hardware Implementation Trade-Offs of Polynomial Approximations and Interpolations," IEEE Transactions on Computers, vol. 57, no. 5, pp. 4093-4104, May 2008.
[38]Xilinx, "The Zynq Book," PDF File, http://www.zynqbook.com/, 2014.
[39]Xilinx, "Designing-IP-Subsystems-Using-IP," PDF File, http://www.xilinx.com/support/documentation/sw_manuals/xilinx2014_1/ug994-vivado-ip-subsystems.pdf, May 2014.
[40]Xilinx, "Accelerating OpenCV Applications with Zynq-7000 All Programmable SoC using Vivado HLS Video Libraries," PDF File, http://www.xilinx.com/support/documentation/application_notes/xapp1167.pdf , June 2015.
[41]Z. 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, pp. 600-612, Apr. 2004.