臺灣博碩士論文加值系統

目前於大型區域安全監控上，隨著監控範圍愈大，所需的攝影機數量跟著增加。整個系統在處理與傳輸影像的負擔也大幅的提升，如何有效解決系統負擔已是目前大型區域安全監控上急需解決的重要課題。本研究朝著一個新的思考方向，嘗試使用日趨成熟的感測網路來幫忙解決大型區域安全監控所遇到的問題。於大型區域中，關鍵區域、重要通道及出入口除了架設攝影機外，放置感測器作為位置感知系統(Location-Aware System)。位置感知系統使用滑動窗口(Slide Window)的觀念，以半秒為一個窗口，從半秒中各種感測資料的變化情況找出代表各類事件的特徵。並以其標準差作正規化來提高鑑別度，經由SVM訓練建造出一個分類器，對未來資料每秒作八次分類判斷，實驗分類正確率約99.34%。將位置感知系統結合攝影機監控，當位置感知系統感測到有事件發生時才去做前景偵測；未有事件發生時僅做背景更新，以提高背景更新之可靠度。不需透過複雜的運算就可有效提升前景偵測的正確率。與一般大型區域安全監控相比，大幅降低系統在處理與傳輸影像上的負擔。本研究使用的視訊偵測系統實際搭配感測器應用後，可提升無事件發生時系統的處理速率。

In general, large amount of cameras are demanded to surveil a large building or area but this may introduces a high data transmission rate and computation cost. The reason why a high data transmission rate is required is that no matter what a event occurs the video bit stream form the cameras are transmitted continuously. In this study, a location-aware system using the sensor network is developed and then cooperated with the vision-based surveillance system to construct a new sensor-based surveillance system. When the location-aware system detects an event, then the object detection scheme is performed in the video surveillance system. Otherwise, only the background updating process works such that the high data transmission rate can be reduced. By extracting the features from the multi-sensor data within a sliding window, the location-aware system uses the SVM classifier to detect the various kinds of events. The experimental results show that the accuracy of event detection may achieve 99.34% for the proposed system.

第一章 序 論．．．．．．．．．．．．．．．．．．．．．．．．．．．1
1.1 研究動機．．．．．．．．．．．．．．．．．．．．．．．．．．1
1.2 相關研究．．．．．．．．．．．．．．．．．．．．．．．．．．2
1.3 系統流程．．．．．．．．．．．．．．．．．．．．．．．．．．6
1.4 論文架構．．．．．．．．．．．．．．．．．．．．．．．．．．9
第二章 感測網路．．．．．．．．．．．．．．．．．．．．．．．．．．10
2.1 無線感測器簡介．．．．．．．．．．．．．．．．．．．．．．．10
2.2 封包格式．．．．．．．．．．．．．．．．．．．．．．．．．．13
第三章 感測器訊號之特徵擷取．．．．．．．．．．．．．．．．．．．．16
3.1 各事件變化圖．．．．．．．．．．．．．．．．．．．．．．．．16
3.2 特徵擷取．．．．．．．．．．．．．．．．．．．．．．．．．．23
第四章 位置感知系統．．．．．．．．．．．．．．．．．．．．．．．．29
4.1 Support Vector Machine．．．．．．．．．．．．．．．．．．．．29
4.1.1 基本觀念．．．．．．．．．．．．．．．．．．．．．．．29
4.1.2 基本原理．．．．．．．．．．．．．．．．．．．．．．．29
4.1.3 核函數．．．．．．．．．．．．．．．．．．．．．．．．31
4.2 SVM訓練．．．．．．．．．．．．．．．．．．．．．．．．．33
第五章 物件偵測．．．．．．．．．．．．．．．．．．．．．．．．．．34
5.1 背景機率模型建立．．．．．．．．．．．．．．．．．．．．．．34
5.2 空間高斯分佈模型．．．．．．．．．．．．．．．．．．．．．．35
5.3 空間-時間統計模型 ．．．．．．．．．．．．．．．．．．．．．36
5.4 背景機率模型更新．．．．．．．．．．．．．．．．．．．．．．37
第六章 實驗結果．．．．．．．．．．．．．．．．．．．．．．．．．．39
6.1 感測器設置．．．．．．．．．．．．．．．．．．．．．．．．．39
6.2 特徵擷取．．．．．．．．．．．．．．．．．．．．．．．．．．40
6.3 SVM訓練 ．．．．．．．．．．．．．．．．．．．．．．．．．43
6.4 事件偵測．．．．．．．．．．．．．．．．．．．．．．．．．．44
6.4.1 未有事件情況．．．．．．．．．．．．．．．．．．．．．44
6.4.2 開門事件情況．．．．．．．．．．．．．．．．．．．．．45
6.4.3 開窗事件情況．．．．．．．．．．．．．．．．．．．．．46
6.4.4 經過事件情況．．．．．．．．．．．．．．．．．．．．．48
第七章 結論與未來工作．．．．．．．．．．．．．．．．．．．．．．．50
7.1 結論．．．．．．．．．．．．．．．．．．．．．．．．．．．．50
7.2 未來工作．．．．．．．．．．．．．．．．．．．．．．．．．．50
參考文獻．．．．．．．．．．．．．．．．．．．．．．．．．．．．．．52

[1] I. F. Akyidiz, W. Su and Y. Sankarasuramniam, “A survey on sensor network,” IEEE Communication Magazine, Vol. 40, No. 8, pp. 102-114, Aug. 2002.
[2] Q. Cao, T. Abdelzaher, T. He and J. Stankovic, “Towards optimal sleep scheduling in sensor networks for rare-event detection,” Fourth International Conference on Information Processing in Sensor Networks (IPSN '05), pp. 20-27, April 2005.
[3] Y. Li, J. Harms and R. Holte, “Optimal traffic-oblivious energy-aware routing For multihop wireless networks,” 25th IEEE International Conference on Computer Communications (INFOCOM 2006), pp. 1-12, April 2006.
[4] M. Elmusrati, R. Jantti and H. Koivo, “Distributed sensor network data fusion using image processing,” 2005 Systems Communications (ICW'05, ICHSN'05, ICMCS'05, SENET'05), pp. 383-388, Aug. 2005.
[5] H. Ma and Y. Liu, “Correlation based video processing in video sensor networks,” 2005 International Conference on Wireless Networks, Communications and Mobile Computing, Vol. 2, pp. 987- 992, June 2005.
[6] Z. He and D. Wu, “Performance analysis of wireless video sensors in video surveillance,”IEEE Global Telecommunications Conference (GLOBECOM '05), Vol. 1, pp. 178-182, Dec. 2005.
[7] S. Soro and W. B. Heinzelman, “On the coverage problem in video-based wireless sensor networks,” 2nd International Conference on Broadband Networks, Vol. 2, pp. 932- 939, Oct. 2005.
[8] A. Prati, R. Vezzani, L. Benini, E. Farella and P. Zappi, “An integrated multi-modal sensor network for video surveillance,” Third ACM international workshop on Video surveillance & sensor networks, pp. 95–102, 2005.
[9] J. R. Morrison, “A line of sight sensor network for wide area video surveillance: simulation and evaluation,” International Conference on Innovative Applications of Distributed Sensor Networks. Bethesda, Oct. 18-19, 2005.
[10] P. Kulkarni, D. Ganesan and P. Shenoy, “The case for multi-tier camera sensor networks,” International Workshop on Network and Operating System Support for Digital Audio and Video (NOSSDAV '05), pp. 141-146, 2005.
[11] A. Elgammal, D. Harwood and L. Davis, “Non-parametric model for background subtraction,” 6th European Conference on Computer Vision, pp. 751-767, 2000.
[12] X. Dai and S. Khorram, “Performance of optical flow techniques,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 236-242, June 1992.
[13] C. C. Lien, J. C. Wang and Y. M. Jiang, “Multi-mode target tracing on s crowd scene,” IEEE International Conference on Intelligent Information Hiding and Multimedia Signal Processing 2007 (IIH-MSP 2007), Nov. 26-28, Kaohsiung, Taiwan. (EI)
[14] TinyOS Community Forum, http://www.tinyos.net/
[15] Crossbow Technology, Wireless Sensor Networks User Manuals, http://www.xbow.com/Support/wUserManuals.aspx
[16] C. J. C. Burges, “A tutorial on support vector machines for pattern recognition,” Journal of Data Mining and Knowledge Discovery, Vol. 2, No. 2, pp. 121-167, 1998.
[17] H. T. Lin and C. J. Lin, “A study on sigmoid kernels for SVM and the training of non-PSD kernels by SMO-type methods,” Technical Report, Department of Computer Science & Information Engineering, National Taiwan University, 2003.
[18] C. C. Chang and C. J. Lin, LIBSVM : A Library for Support Vector Machines, http://www.csie.ntu.edu.tw/~cjlin/libsvm/index.html
[19] C. W. Hsu, C. C. Chang, and C. J. Lin, “A practical guide to support vector classification,” Technical Report, Department of Computer Science & Information Engineering, National Taiwan University, Taiwan, 2003.
[20] Y. Ren, C. S. Chua and Y. K. Ho, “Motion detection with nonstationary background,” Machine Vision and Applications, Vol. 13, No. 5-6, pp. 332-343, Mar. 2003.
[21] C. C. Lien and S. C. Hsu, “The target tracking using the spatial-temporal probability model,” IEEE International Workshop on Nonlinear Signal and Image Processing, pp. 393-398, May 18-20 2005, Sapporo, Japan.
[22] S. Tulyakov and V. Govindaraju, “Identification model with independent matching scores”, Biometrics Consortium Conference, Crystal City, VA, Sep. 2005.