物件追蹤(Object Tracking)在視訊影像處理方面上一直是個重要的議題，透過物件的追蹤，可以有效了解到特定區域中物體的動作與去向，監視系統(Surveillance System)由於必須長時間監視並紀錄，基於在成本上的考量，大多數的使用者選擇低解析度(Low Resolution)的監視系統，因此當需要錄製影像提供資訊時，常常因為解析度太低導致辨識上的困難。本研究針對這一個問題，提出一個在監視系統視訊上快速處理追蹤物件的演算法，並且能將擷取之影像加以細部強化以改進後續辨識上的成功率。

Due to the cost consideration, most of surveillance systems adopt the low resolution format to record the video sequences. However, the low resolution image quality often results the interested object too vague to be identified. We propose a two folds algorithm- tracking and enhancement- to solve this problem. First, the block-based motion estimation is used for object tracking. Next, a novel image enhancement scheme is used to reconstruct an initial high resolution image acquired from the region of interested (ROI). The error back-projection is further used to improve the quality of the reconstructed high resolution image. The proposed algorithm has been tested on surveillance video sequences and some common video sequences. The tracking results are satisfying that it demonstrated the proposed algorithm is illumination invariant and robust to complex background. The image enhancement scheme is also tested. The test results on synthetic images showed that the quality of enlarged images has been improved.

目錄
淡江大學論文提要	I
英文摘要	II
目錄	III
圖目錄	IV
表目錄	VI
第一章 緒論	1
1.1研究動機與目的	1
1.2系統流程	2
1.3 章節組織	4
第二章 相關研究	5
2.1 物件追蹤	5
2.1.1以區塊為基礎的動量估計	5
2.1.2 宏塊(Macroblock)	6
2.1.3 動量估計演算法	7
2.1.4 採樣(Subsampling)	21
2.2 影像強化	23
2.2.1反投影迭代法(Iterative Back-Projection, IBP)	23
第三章 提出的方法	25
3.1物件追蹤	25
3.2影像強化	29
3.2.1初始影像重建	29
3.2.2誤差反投影 (Error Back-Projection)	34
第四章 實驗結果與分析	38
4.1 物件追蹤之實驗結果	38
4.2影像強化	42
4.2.1初始影像重建的演算法	42
4.2.2誤差補償時的權重	42
4.2.3 EBP的效能	45
第五章 結論與未來研究	47
參考文獻	48
附錄-英文論文	50
圖目錄
圖1物件追蹤流程圖	2
圖2影像強化流程圖	3
圖3以區塊為基礎的動量估計示意圖	6
圖4 Full-Search 演算法示意圖	8
圖5 TSS演算法中第一回合的測量基準點	9
圖6 TSS演算法的運作過程，其中●為第一回合的測量點，■為第二回合的測量點，◆為第三回合的測量點。	10
圖7使用FA演算法反覆計算100次測量出的動量分佈情形[7]	11
圖8 NTSS演算法流程圖	12
圖9 NTSS演算法中第一回合17個測量基準點	13
圖10 NTSS演算法中若第一回合結果為中心8 鄰居點時，第二回合額外計算點	14
圖11 4SS演算法計算之基準點	15
圖12 4SS演算法中兩種不同的搜尋路徑範例[8]	16
圖13 ITSS演算法的兩種不同搜尋路徑範例[10]	17
圖14 DS演算法所使用之兩種搜尋模式	18
圖15 DS演算法計算上可能之情形(CASES)	19
圖16 DS演算法搜尋路徑範例圖[12]，圖中最終動量為(-4, 2)	20
圖17 OLS演算法示意圖。●為第一回合、▂為第二回合、▲、◆分別為第三與第四回合、■則為最終回合	21
圖18像素匹配使用模組[11]	22
圖19本論文提出之動量估算演算法初始測量基準點	26
圖20本論文提出之追蹤演算法流程圖	27
圖21本論文提出之動量分佈情形決策方式	28
圖22初始影像重建演算法流程圖	31
圖23放大兩倍的影像效果比較	32
圖24 Barbara重建的過程	33
圖25 EBP演算法流程圖[16]	36
圖26執行完EBP後的結果	37
圖27室內視訊影像之人臉追蹤(紅框為追蹤之區域)	39
圖28室內視訊影像且有強烈光源之實驗結果	39
圖29室外視訊影像場景追蹤實驗結果	40
圖30複雜背影之視訊影像追蹤	41
圖31 Lena測試影像在不同權重係數下PSNR值折線圖	44
圖32 Barbara測試影像在不同權重係數下PSNR值折線圖	45
圖33 Lena與Barbara兩張測試影像重建以及誤差反投影演算法之實驗結果	46
圖34實際影像經由提出的方法計算後之結果。圖下數字為該圖之解析度	46
表目錄
表1  bicubic之影像與本論文提出之影像重建方法PSNR值	42
表2  Lena 測試影像在不同權重係數下PSNR值	43
表3  Barbara 測試影像在不同權重係數下PSNR值	44

參考文獻
1.	L. Li, W. Huang, Y.H Gu, and Q. Tian, "Statistical Modeling of Complex Backgrounds for Foreground Object Detection, "IEEE Trans. image processing, Vol 13, pp.1459 - 1472, Nov. 2004. 
2.	J.W. Hsieh, S.H. Yu, and W.F. Hu "Automatic Traffic Surveillance System for Vehicle Tracking and Classification, "IEEE Trans. Intelligent Transportation Systems, Vol 7, pp.175 - 187, June 2006.
3.	B. Lei and L.Q. Xu, "Real-time outdoor video surveillance with robust foreground extraction and object tracking via multi-state transition management, "Pattern Recognition, Vol.27, No.15, pp.1816-1825, Nov 2006.
4.	Y.M. Mustafah, T.Shan, A. W. Azman, A. Bigdeli,and B.C.Lovell "Real-Time Face Detection and Tracking for High Resolution Smart Camera System, "Digital Image Computing Techniques and Applications, pp.387 - 393, Dec. 2007.
5.	J. Cui, H. Zha, H. Zhao, and R. Shivasaki, “Multi-modal tracking of people using laser scanners and video camera, ” Image and Vision Computing, Vol.26, No.2, pp. 240-252, June 2008.
6.	T. Koga, K. Iinuma, A. Hirano, Y. Iijima, and T. Ishiguro, “Motion-compensated interframe coding for video conferencing,” in Proc. NTC 81, pp. C9.6.1-9.6.5, New Orleans, LA, Nov./Dec. 1981.
7.	R. Li, B. Zeng, and M. L. Liou, “A new three-step search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Technol., vol. 4, pp. 438- 442, Aug. 1994.
8.	L. M. Po and W. C. Ma, “A novel four-step search algorithm for fast block motion estimation,” IEEE Trans. Circuits Syst. Video Technol, vol. 6, pp.313-317, June 1996
9.	L. K. Liu and E. Feig, “A block-based gradient descent search algorithm for block motion estimation in video coding,” IEEE Trans. Circuits Syst. Video Technol, vol. 6, pp. 419-422, Aug. 1996.
10.	Donglai Xu, C. Bailey and R. Sotudeh, "An Improved Three-Step Search Block-Matching Algorithm for Low Bit-Rate Video Coding Applications," Signals, Systems, and Electronics, ISSSE 98. pp. 178 - 181, Oct. 1998
11.	F.H. Cheng, S.N. Sun, New Fast and Efficient Two-Step Search Algorithm for Block Motion Estimation, "IEEE Trans. Circuits Syst. Video Technol, Vol 9,pp. 977-983, Oct. 1999
12.	S.Zhu and K. K. Ma, “A new diamond search algorithm for fast block-matching motion estimation,” IEEE Trans. Image Processing, vol.9, pp. 287-290, Feb.2000.
13.	C. Zhu, X. Lin, and L. P. Chau, “Hexagon-based search pattern for fast block motion estimation,” IEEE Trans. Circuits Syst. Video Technol, vol. 12, pp. 349-355, May 2002.
14.	Xuan Jing and Lap-Pui Chau, “An efficient three-step search algorithm for block motion estimation,” IEEE Trans. Multimedia, vol.6, pp. 435-438, JUNE 2004.
15.	S. Soongsathitanon, W.L. Woo, and S.S. Dlay, "Fast Search Algorithms for Video Coding using Orthogonal Logarithmic Search Algorithm," IEEE Trans. Consumer Electuonics, Vol. 51, pp.552-559, MAY 2005
16.	J.S. Park, S.W. Lee, "Resolution enhancement of facial image using an error back-projection of example-based learning, " Sixth IEEE International Conference Automatic Face and Gesture Recognition, pp.831-836, May 2004
17.	G.S. Lin, and M.K. Lai "Enhancing Resolution Using Iterative Back-Projection Technique for Image Sequences, "Journal of Computers, Vol.19, Nov 2008