 |
|
|
|
以作者查詢圖書館館藏 、以作者查詢臺灣博碩士 、以作者查詢全國書目 、勘誤回報 、線上人數：74 、訪客IP：3.142.135.186
姓名 陳俊樺(Chun-hua Chen) 查詢紙本館藏 畢業系所 電機工程學系 論文名稱 基於高階幾何特徵及卡爾曼協同追蹤之監控系統
(Smart Surveillance System Based on High Level Features and Co-tracking by Kalman Filter)相關論文 檔案 [Endnote RIS 格式]
[相關文章]
[文章引用]
[完整記錄]
[館藏目錄]
[檢視]
- 本電子論文使用權限為同意立即開放。
- 已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
- 請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。
摘要(中) 隨著時代的進步，安全也隨著被大家所重視。然而在現今社會裡，自動化與智慧型的系統也成為目前科技所追求的。監控系統的環境可應用在各個環境下，例如機場，高速公路以及校園等等…。在這些環境中，常會有複雜的情形發生，例如多人交錯、多人並肩走在一起或者人的靜止，這些情況在追蹤都會是一個很嚴重的問題。因此如何有效解決以上所提的三大問題在追蹤這方面一直是很重要的議題。
在此論文中，為了解決含有以上所述的一些問題，擷取重要的幾何特徵就成為了達到精確的追蹤系統的主要方法之一。擷取幾何特徵的方法運用一個簡單且直覺的影像處理所獲得。然而幾何擷取特徵的特性含有唯一性及具有意義的。另外，在追蹤的系統上，運用這兩個重要的幾何特徵，並行的使用追蹤演算法，增加追蹤的準確度。最後在靜止物件上，同樣使用了這樣的幾何特徵，解決靜止物件所產生的相關問題，減少物件遺失的情形。實驗顯示提出的方法在上述的三個問題都能有好的結果。除此之外，為了顯示提出方法有效的改善問題，將結果轉為數據，並與另外兩個方法做比較。實驗的影片也採用被採信的基準影片PETS2001、PETS2006和PETS2007。
除了軟體上的開發，也將其實現在DM642開發平台上，此系統利用一監視器傳輸到DM642開發板上進行追蹤處理，再經由液晶螢幕顯示出來。在DM642平台上跑了一實例來驗證結果，並評估其效能。
摘要(英) Tracking multiple targets in complex situation is challenging. The difficulties are tackled multiple targets with occlusions, especially when multiple involved targets are grouped, moving together in appearance ,and target is static. These problems are still focused in recent years.
In this paper, we present a multiple targets tracking system for the management of occlusion problem. The proposed algorithm introduces a geometric shape co-tracking strategy. It decomposes targets into geometric shapes located on body and head parts based on reasonable target geometry consideration. Features selected from the decomposed geometric shapes then can be used to track targets through intersections such as occlusion. Projection histogram and ellipsoid shape model are adopted to manage decomposed geometric shapes corresponding to each target. Tracking is done through Kalman filtering process with high efficient and low complexity issue. The problem of sleeping object is handled by mean shift with no foreground needed. Experimental results show that the occlusion of grouped targets can be tracked successfully on recent challenging benchmark sequences.
The proposed method is also implemented in DM642. The input video is from a camera transfer to DM642, then the input video has been through tracking process and output the video in LCD. To verify the system, a realistic video has been shown also.
關鍵字(中) ★ 協同追蹤
★ 監控系統
★ 幾何特徵關鍵字(英) ★ surveillance system
★ co-tracking
★ geometric feature論文目次 摘要 I
ABSTRACT II
誌謝 III
目錄 IV
圖目錄 VI
表目錄 VIIII
第一章 緒論 1
1.1研究動機 2
1.2論文組織架構 4
第二章 追蹤系統之背景及其相關研究 5
2.1 物件表示法 6
2.2 特徵選擇 8
2.3 概論:物件追蹤 9
2.3.1點追蹤(point tracking) 10
2.3.1.1 決定型的方法 10
2.3.1.2 機率型的方法 11
2.3.2 核心追蹤(kernel tracking) 12
2.3.2.1 模板與密度型態(template and density-based appearance)模組 13
2.3.2.2 多視角型態(multiview based appearance)模組 14
2.3.3 剪影追蹤(silhouette tracking) 15
2.3.3.1 形狀配對(shape matching) 15
2.3.3.2 輪廓追蹤(contour tracking) 16
第三章 型態分解演算法 17
3.1 型態分解之介紹 18
3.2 型態分解之基本定義 19
3.3 型態分解之元素選擇 21
3.4 型態分解 24
第四章 高階特徵及卡爾曼協同追蹤之監控系統 26
4.1系統概論 27
4.2幾何切割 29
4.3 特徵擷取 31
4.4 幾何配對 33
4.5 協同追蹤 35
4.6 靜止物件之處理 37
4.6.1 靜止與移動物件之判斷 37
4.6.2 靜止物件之追蹤 37
4.7 DSP實現及其優化 40
第五章 實驗結果 43
5.1 提出演算法之實驗結果 44
5.1.1 PETS2001 45
5.1.2 PETS2006 48
5.1.3 PETS2007 51
5.2 有無靜止物件之實驗結果 54
5.2.1 PETS2006 54
5.2.2 SC 55
5.3 DSP實驗結果及數據 57
5.3.1 DSP之實驗結果 57
5.3.2 DSP之實驗數據 58
5.4 實驗結果總結 60
第六章 結論 61
REFERENCE 63
參考文獻 [1] A. Yilmaz, O. Javed and M. Shah, "Object Tracking: A Survey", ACM Journal of Computing Surveys, Dec 2006.
[2] Comaniciu, D., Ramesh, V., and Meer, P. “Kernel-based Object Tracking”, IEEE Trnas. Pattern Anal. Machine Intell., 2003.
[3] Isard, M. and MacCormick, J. “Bramble: A Bayesian Multiple-blob Tracker”, IEEE International Conference on Computer Vision, 2001.
[4] Black, M. and Jepson, A. “Eigentracking: Robust Matching and Tracking of Articulated Objects Using A View-based representation” Int. J. Comput. Vision, 1998
[5] Avidan, S. “Support Vector Tracking”, IEEE Conf. on Computer Vision and Pattern Recognition, 2005.
[6] Kang, J., Cohen, I., and Medioni, G., “Object Reacquisition Using Geometric Invariant Appearance Model” International Conference on Pattern Recognition, 2004.
[7] Chen, Y., Rui, Y., and Huang, T. “JPDAF based HMM for Real-time Contour Tracking” IEEE Conf. on Computer Vision and Pattern Recognition, 2001.
[8] J. Xu, “Decomposition of convex polygonal morphological structuring elements into neighborhood subsets”, IEEE Trans. Pattern Anal. Machine Intell., vol. 13, pp. 153–162, Feb. 1991.
[9] J. Xu, “Morphological decomposition of 2-D binary shapes into conditionally maximal convex polygons” , IEEE Trans. Pattern Anal. Machine Intell., 1991.
[10] Texas Instruments, “Code Composer Studio User’s Guide,” Literature number: SPRU328B, February 2000.
[11] Texas Instruments, “TMS320C6000 Optimizing Compiler User’s Guide,” Literature number: SPRU187L, May 2004.
[12] Texas Instruments, “TMS320C64x+ DSP Cache User's Guide,” Literature number: SPRU862B, February 2009.
[13] P. Kumar, S. Ranganath, K. Sengupta, and H. Weimin, “Cooperative Multitarget tracking with efficient split and merge Handling”, IEEE Trans. Circuits and systems for video technology, 2006.
[14] H. Liu, L. Zhang, Z. Yu, H. Zha, Y. Shi, “Collaborative Mean Shift Tracking Based on Multi-Cue Integration and Auxiliary Objects”, IEEE trans. Signal processing, 2007
[15]Smith, K., Gatica-Perez, D., Odobez, J.-M. and Ba, S., “Evaluating Multi-object Tracking”, IEEE Conf. on Computer Vision and Pattern Recognition, 2005.
[16] T.-H. Tsai, W.-T. Sheu and C.-Y. Lin, "Foreground Object Detection based on Multi-model Background Maintenance ", IEEE International Symposium on Multimedia, ISM ’07, Taiwan , 2007.
指導教授 蔡宗漢(Tsung-Han Tsai) 審核日期 2010-3-19 推文 plurk
funp
live
udn
HD
myshare
netvibes
friend
youpush
delicious
baidu
網路書籤 Google bookmarks
del.icio.us
hemidemi
facebook
twitter
google
reddit