OpenVX是一個開放、免授權費的跨平台標準，可用來加速嵌入式系統電腦視覺應用。它能實現效能與功耗最優化的電腦視覺處理，對臉部、身體和手勢追蹤、智慧視訊監控、先進駕駛輔助系統(ADAS)、擴增實境等應用在嵌入式系統的real-time處理來說特別重要。本論文深入探討OpenVX的內容，在OpenVX的標準下，開發了一套人臉動作模擬系統。其中針對OpenVX API之亮度轉換，提出專有硬體架構，並在FPGA實驗板實作嵌入式軟硬體協同運算。如此系統處理速度能進一步提升，且只需增加少量硬體成本。

OpenVX, an open, royalty-free, cross-platform standard, can be used to speed up computer vision applications in embedded systems. It can achieve performance and power-optimized processing of computer vision, including facial, body and gesture tracking, intelligent video surveillance, advanced driver assistance systems (ADAS) and augmented reality on real-time embedded systems. In this thesis, we investigates the contents of OpenVX specification and develop an OpenVX based computer vision algorithm for face detection and gesture tracking. FPGA implementation of hardware and software co-operation is also demonstrated where the color-to-intensity conversion is executed in hardware. It is shown that the performance of the system can be enhanced without increasing too much hardware costs.

論文審定書 i
中文摘要 ii
Abstract iii
目錄 iv
圖目錄 viii
表目錄 x
第一章、概論 1
1.1　本文大綱 1
1.2　研究動機 1
第二章、研究背景與相關研究 3
2.1　OpenVX 3
2.2　臉部特徵偵測 6
2.3　人臉動作模擬 9
2.4　相關研究總結 10
第三章、OpenVX標準 11
3.1　概述 11
3.1.1　命名規則 11
3.1.2　專業術語和首字母縮略詞 11
3.2　OpenVX標準中的物件 12
3.2.1　框架物件 12
3.2.2　資料物件 18
3.2.3　錯誤物件 21
3.3　與Graph有關的概念 22
3.3.1　連接Node 22
3.3.2　虛擬資料物件 22
3.3.3　Node參數 23
3.3.4　執行模型 23
3.3.5　驗證 24
3.4　基礎視覺函式 24
3.4.1　Channel Extract 26
3.4.2　Color Convert 26
3.5　生命週期 27
3.5.1　Context的生命週期 27
3.5.2　Graph的生命週期 27
3.6　OpenVX的程式範例 28
3.7　程式設計環境的搭建 29
3.7.1　CMAKE 29
3.7.2　Concerto 31
第四章、利用OpenVX API實現電腦視覺應用 33
4.1　實驗環境 33
4.2　臉部特徵偵測 33
4.2.1　光線補償 33
4.2.2　人臉偵測 34
4.2.3　人臉眼睛部份偵測 35
4.2.4　定位眼睛、嘴巴特徵點位置 36
4.3　人臉動作模擬 37
4.3.1　高斯模糊金字塔 37
4.3.2　光流法追蹤特徵點 37
4.3.3　紀錄移動向量 38
4.3.3　將移動向量套入測試者 38
4.4　在OpenVX標準下的演算法流程 39
第五章、OpenVX軟硬體協同設計 44
5.1　實驗環境 44
5.2　硬體架構設計 44
5.2.1　AXI Protocol 44
5.2.2　硬體架構說明 45
5.2.3　硬體波型說明 47
5.3　軟硬體協同運算架構 48
5.3.1　在Zedboard上啟動Linux作業系統 49
5.3.2　加入 DMA Engine 50
5.3.3　加入硬體加速器 52
5.3.4　軟體應用程式協同硬體運算 52
第六章、實驗結果及分析 55
6.1　臉部特徵偵測結果 55
6.2　人臉動作模擬結果 56
6.3　軟體執行耗時分析 57
6.3.1　OpenVX Grap內的執行時間分析 57
6.3.2　OpenVX Graph外的執行時間分析 58
6.3.3　OpenVX Graph內 vs OpenVX Graph外之分析 60
6.4　軟硬體協同運算執行耗時分析 61
6.4.1　 OpenVX Graph內的執行時間分析 61
6.4.2　硬體執行效率分析 62
6.4.3　總程式執行時間分析 62
第七章、結論與未來展望 64
7.1　結論 64
7.2　未來展望 64
參考文獻 65

[1] S. Gautham and E. Rainey. The Khronos OpenVXTM1.0 Specification.The Khronos Group, 2014.
[2] OpenVX Ecosystem Overview. OpenVX Workshop, Embedded Vision Summit, May 2016
[3] Rainey, Erik, et al. "Addressing System-Level Optimization with OpenVX Graphs." Computer Vision and Pattern Recognition Workshops (CVPRW), 2014 IEEE Conference on. IEEE, 2014.
[4] Tagliavini, Giuseppe, Germain Haugou, and Luca Benini. "Optimizing memory bandwidth in OpenVX graph execution on embedded many-core accelerators." Design and Architectures for Signal and Image Processing (DASIP), 2014 Conference on. IEEE, 2014
[5] 黃孟隆(2004)。在測謊作業上之瞳孔反應(未出版之碩士論文)。中央警察大學， 桃園縣。
[6] Ueno, H., Kaneda, M., & Tsukino, M. (1994). Development of drowsiness detection system. in Proc. Conf. Vehicle Navigation and Information Systems ,Yokohama, Japan, 15-20.
[7] K. Grauman, M. Betke, J. Gips, and G. R. Bradski.( 2001).Communication via Eye Blinks-Detection and Duration Analysis in Real Time. IEEE International Conference on Computer Society, 1010-1017.
[8] T. Lee, S. K. Park, and Mignon Park,(2005). A New Facial Features and Face Detection Method for Human-Robot Interation. Proceedings of the 2005 IEEE International Conference on Robotics and Automation, 2063-2068.
[9] W. Dong and X. Wu.(2005).Driver Fatigue Detection Based on the Distance of Eyelid. IEEE International Workshop on VLSI Design & Video Tech, 365-368.
[10] 黃泰祥(2000)。具備人臉追蹤與辨識功能的一個智慧型數位監視系統(未出版之 碩士論文)。中原大學，桃園縣。
[11] M. Soriano, B. Martinkauppi, S. Huovinen, and M. Laaksonen.(2000). Using the Skin Locus to Cope with Changing Illumination Conditions in Color-Based Face Tracking.IEEE Symposium on Signal Processing, 383-386.
[12] W. Rongben, G. Lie, T. Bingliang, and J. Lisheng.(2004).Monitoring Mouth Movenent for Driver Fatigue or Distraction with One Camera. IEEE 81 International Conference on Intelligent Transportation Systems, 314-316.
[13] P. Smith, M. Shah and N. d. V. Lobo.(2004).Determining Driver Visual Attention with One Camera. IEEE Transactions on Intelligent Transportation Systems, 4, 205-218. [18] Y. Yacoob and L. D. Davis “Recognizing human facial expressions from long image sequences using optical flow.” IEEE Trans.on Pattern Analysis and Machine Intelligence, vol. 18, no. 6, pp.636-642, 1996.
[14] Y. Yacoob and L. D. Davis “Recognizing human facial expressions from long image sequences using optical flow.” IEEE Trans.on Pattern Analysis and Machine Intelligence, vol. 18, no. 6, pp.636-642, 1996.
[15] J. Lien, T. Kanade, J. Cohn, and C. Li, “Detection, tracking, and classification of action units in facial expression,” Robotics andAutonomous Systems, vol. 31, Issue: 3, pp. 131-146, May 2000.
[16] M. S. Bartlett, J. C. Hager, P. Ekman, and T. J. Sejnowski,“Measuring Facial Expressions by Computer Image Analysis,”Psychophysiology, vol. 36, pp. 253-263, 1999.
[17] P. Ekman and W.V. Friesen, The Facial Action Coding System: A Technique for The Measurement of Facial Movement. San Francisco: Consulting Psychologists Press, 1978.
[18] 潘奕安，「低解析度影像序列知自動化表情辨識系統」，成功大學資訊工程所，2004。
[19] 何明哲，「以模糊推論進行臉部動作元之分析與辨識」，東華大學資訊工程所，2004。
[20] J. F. Cohn, A. J. Zlochower, J. J. Lien, and T. Kanade,“Feature-point tracking by optical flow discriminates subtle differences in facial expression,” in Proceedings of the 3rd IEEE International Conference on Automatic Face and Gesture Recognition, pp. 396–401, April 1998.
[21] Hsu, R. L., A. M. Mohamed and A. K. Jain (2002) Face detection in color image. IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(5), 696-704.
[22] Soriano, M., B. Martinkauppi, S. Huovinen and M.Laaksonen (2000) Skin color modeling under varying illumination conditions using the skin locus for selecting training pixels. Proceedings of IEEE Nordic Signal Processing Symposium, Kolmarden, Sweden.
[23] Soriano, M., B. Martinkauppi, S. Huovinen and M.Laaksonen (2000) Skin detection in video under changing illumination conditions. Proceedings of the 15th International Conference on Pattern Recognition,Barcelona, Spain.
[24] B.D. Lucas and T. Kanade, “An Iterative Image Registration Technique with an Application to Stereo Vision”, DARPA Image Understanding Workshop, pp. 121-130, 1981.(see also IJCAI’81, pp. 674-679)