臺灣博碩士論文加值系統

近年來社會犯罪案件頻繁，銀行、超商及住家…等的搶案不斷發生，為能減低及嚇阻犯罪案件的增加，監控系統的存在，就彰顯出其極其重要性。而拜科技進步所賜，監視器材來的比以往更為精密及便宜，其也使得監控保全方面的系統能以大量的開發。然而，極大部份的監控系統仍然是以靜態的方式來進行固定視角的定點偵測，但是這樣的方發卻有著監測死角的問題；相反的，為能進行更寬廣的監控範圍，其則必須加裝更多的監視器，以求能解決監控死角的問題。只是，成本需求必然可觀。
也因此，為能解決固定監控視角及成本兩者問題，本論文提出一套結合PTZ (Pan/Tilt/Zoom, PTZ)攝影機的智慧型自動監控保全系統。在本系統中，則是利用PTZ攝影機拍攝所需監測的局部環境影像，其次，利用影像接合技術，以合併起PTZ攝影機所拍攝的各個角度的影像，之後，則是針對接合起的全景影像來進行相關的特徵資訊分析。而為了能分辨出監測環境中是否有存在異常物體，系統則是透過CBIR技術來進行即時影像與全景影像之間的比對，藉由比對的結果來判斷PTZ攝影機目前所在的監控位置，並藉由影像相減來得到兩張影像之間的差別。一旦系統取得異常物體，則會立刻對異常物體進行追蹤程序或者辨識之用。

The social crime case is frequent in recent years, the robing cases of bank, ultra trader and household, etc. are happenning constantly, in order to lower and frighten the increase which hinder the crime case, the existence of the monitoring system, show its importance deeply clearly. And visit the scientific and technological progress to grant, monitor whom apparatus come than in the past more accurate and cheaper, it make it save from damage system of can be with a large amount of development not to control too. But, extremely most monitoring systems remain and carry on detecting examining by fixed position of regular visual angle by way of static behavior, but such a square hair has a question of monitoring the dead angle; on the contrary, in order to carry on more broad control range, its must install more monitors additional, in order to solve the problem that control the dead angle. Just, the demand of the cost must be considerable.
In order to solve and control the visual angle and the two problem of cost regularly, this thesis proposes one set accords with PTZ (Pan/Tilt/Zoom, PTZ) Controlling the system of saving intelligently from damage automatically of the camera. In this system, utilize PTZ camera to shoot local environment images needed to monitor, secondly, utilize joint technology of the image, in order to amalgamate the image of all angles that PTZ camera shoots, later, it was the panoramic image of getting up to the joint that carried on relevant characteristic information analysis. And in order to say whether to have unusual object in the environment of monitoring or not, the system is through right compared with the panoramic image that CBIR technology carries on the instant image, than right result come in control position, and reduce and get the difference between two images with the image. Obtain unusual object systematically, can is it use to go on procedure of tracking or distinguish to unusual object at once.

摘要 I
ABSTRACT II
誌 謝 III
目 錄 IV
圖目錄 VII
第1章 緒論 1
1.1 研究動機 1
1.2 研究目的 2
1.3 研究範圍與限制 3
1.4 論文架構 3
第2章 相關研究 5
2.1 影像接合相關研究 5
2.2 以影像內容之影像檢索相關研究 6
2.3 PTZ監控系統相關研究 7
第3章 系統架構 9
3.1 系統概要 9
3.2 系統架構說明 13
3.2.1 特徵點偵測模組 14
3.2.2 影像接合模組 15
3.2.3 異常比對模組 16
3.2.4 特徵參數設定模組 17
3.2.5 影像前置處理設定模組 18
3.2.6 PTZ攝影機控制模組 18
3.2.7 資料庫模組 19
3.2.8 瀏覽/檢視模組 20
第4章 特徵點偵測系統 21
4.1 影像特徵點偵測相關技術 21
4.2 MORAVEC CORNER DETECTION 22
4.3 HARRIS CORNER DETECTION 24
第5章 影像接合系統 29
5.1 影像接合相關技術 29
5.2 特徵匹配 30
5.3 SIFT(SCALE INVARIANT FEATURE TRANSFORM) 32
5.4 RANSAC(RANDOM SAMPLE CONSENSUS) 38
第6章 異常比對系統 41
6.1 特徵擷取 41
6.1.1 顏色特徵 41
6.1.2 特徵區塊顏色統計 42
6.1.3 物體模型與核心函數 43
6.1.4 形狀特徵 45
6.2 特徵相似性比對 48
6.2.1 形狀比對 48
6.2.2 物體模型比對 48
第7章 實驗成果 51
7.1 系統環境 51
7.2 實驗結果 51
7.3 影像接合 56
7.4 特徵偵測 62
7.5 特徵參數設定介面 64
7.6 影像處理設定介面 64
7.7 影像比對 66
第8章 結論 69
8.1 目前成果 69
8.2 未來展望 70
參考文獻 72

[1]R. D. T. Janssen, and A. M. Vossepoel, “Compilation of mosaics from separately scanned line drawings”, Proceedings of the Second IEEE Workshop on Appliactions of Computer Vision, Netherlands., pp.36-43, 1994.
[2]J. Flusser and T. Suk, “A moment-based approach to registration of images with affine geometric distortion”, IEEE Trans. On Geoscience and Remote Sensing., Vol. 32, No. 2, March, pp. 382-387, 1994.
[3]H. Li, B. S. Manjunath, and S. K. Mitra, “A contour-based approach to multisensor image registration”, IEEE Trans. Image Processing, Vol. 4, NO. 3, pp.320-334, March 1995.
[4]J. W. Hsieh, H. Y. M. Liao, and Y. P. Hung, “Image registration using a new edge-based approach”, 8th IPPR Conference on Computer Vision, Graphics and Image Processing, pp. 20-23, Aug. 1995.
[5]C. W. Lin., C. M. Wang, Y. J. Chang, and Y. C. Chen, “Realtime object extraction and tracking with an active camera using image mosaics”, IEEE Workshop on Multimedia Signal Processing, pp. 149-152, 2002.
[6]Z. P. Liang, H. Pan, R. L. Magin, N. Ahuja, and T. S. Huang, “Automated image registration by maximization of a region similarity metric”, Image Processing, 1997, Proceedings, International Conference, Vol. 3, pp. 272-275, 1997.
[7]M. Hirakawa, K. Uchida, and A. Yoshitaka, “Content-based video retrieval using mosaic images”, Proc. IEEE First international Symposium on Cyber Worlds (CW, 02), pp. 161-167, Nov. 06-08, 2002.
[8]M. C. Chung, “The design of image registration system using the feature-based relational graph”, Master Thesis of Dept. Computer Science and Information Engineering, Chung Hua University, Hsin-Chu, Taiwan, R.O.C., July 2003.
[9]J. Flusser and T. Suk, “Affine moment invariants: a new tool for character recognition”, Pattern Recognition Latters, Vol. 4, pp. 433-436, April 1994.
[10]Z. Q. Hong, “Algebraic feature extraction of image for recognition”, Pattern Recognition, Vol. 24, NO. 3, pp. 211-219, 1990.
[11]M. Flickner, H. Swahney, W. Niblack, J. Shley, Q. Huang, B. Dom, M. Gorkani, J. Hafner, D. Lee, D. Petkovic, D. Steele, P. Yanker, “Query by image and video content : The QBIC system”, IEEE Computer 1995.
[12]B. Mehtre, M. Kankanhalli, W. Lee, “Shape measures for conten based image retrieval : A comparison”, Information Processing Management 33 319-337 1997.
[13]J. Cullen, J. Hull, P. Hart, “Document image database retrieval and browsing using texture analysis”, In:Proc. 4th Int. Conf. Document Analysis and Recognition., 718-721 1997.
[14]Alessandro Bevilacqua, Luigi Di Stefano and Pietro Azzari “An Effective Real-Time Mosaicing Algorithm Apt to Detect Motion Through Bacckground Subtraction Using a PTZ Camera”, ARCES-DEIS(Department of Electronics, Computer Science and Systems, IEEE 2005.
[15]Alessandro Bevilacqua and Pietro Azzari “High-quality real time motion detection using PTZ cameras”, ARCES-DEIS(Department of Electronics, Computer Science and Systems, IEEE 2006.
[16]Kiran S. Bhat, Mahesh Saptharishi and Pradeep K.Khosla, “Motion Detection and Segmentation Using Image Mosaics”, 1. Robotics Institute; 2. Electrical and Computer Engineering, Carnegie Mellon University Pittsburgh, PA 15213, IEEE 2000.
[17]許勝政, 連振昌, “於動態背景下發展以馬賽克全景影像為基礎之即時物件追蹤技術”, 中華大學資訊工程研究所.
[18]Jacob Benesty, Dennis R. Morgan, and Juan H. Cho, ”A new class of doubletalk detectors based on cross-correlation”, IEEE Transactions on Speech and Audio Processing, Vol. 8, pp. 168-172, March 2000.
[19]B. H. Friemel, L. N. Boho and G. E. Trahey, “Relative performance of two-dimensional speckle-tracking techniques : normalized correlation, non-normalized correlation and sum-absolute-difference”, IEEE Confs. Vol. 2, pp. 1481-1484, 1995.
[20]K. Briechle and U. D. Hanebeck, “Self-localization of a mobile robot using fast normalized cross correlation”, IEEE SMC ’99 Confernce, Vol. 4, pp. 720-725, 1999.
[21]John G. Proakis, “Digital Communications”, New York: McGraw-Hill, 2001.
[22]M. A. Fischler and R. C. Bolles, “Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography”, Communication Association and Computing Maching, Vol. 24, NO. 6, pp. 381-395, 1981.
[23]O. Chum, J. Matas and J. Kittler, “Locally optimized RANSC”, In Proc. DAGM, Springer-Verlag, 2003.
[24]D. Nister, “Preemptive RANSAC for live structure and motion estimation”, In Proc. ICCV03, pp. 199-206, October 2003.
[25]D. G. David, “Object recognition from local scale-invariant features”, International Conference on Computer Vision, Corfu, pp. 1150-1157, Scptember 1999.
[26]D. G. David, “Distinctive image features from scale-invariant keypoints”, International Journal of Computer Vision, pp. 91-110, 2004.
[27]T. Lindeberg, “Scale-space theory: A basic tool for analysing structures at different scales,” Journal of Applied Statistics, vol. 21, no. 2, pp. 224-270, 1994.
[28]H. Moravec, “Obstacle avoidance and navigation in the real world by a seeing robot rover”, Technical Report CMU-RI-TR-3, Carnegie-Mellon University, Robotics Institute, 1980.
[29]S. M. Smith and M. Brady, “SUSAN – a new approach to low level image processing”, International Journal of Computer Vision, Vol. 23(1), pp. 45-78, 1997.
[30]C. Harris and M. J. Stephens, “A combined corner and edge detector”, In Alvey Vision Conference, pp. 147-152, 1988.
[31]D. Comaniciu, V. Ramesh, and P. Meer, “Kernel Based Object Tracking,” IEEE Trans. Pattern Anal. Machine Intel., Vol. 25, No.5, pp. 564-557, May 2003.
[32]R. M. Haralick, K. Shanmugam and I. Dinstein, “Textural Features for Image Classification”, IEEE Trans. pp. 610-621, 1973.
[33]R. M. Haralick, “Statistical and structural approaches to texture”, Proc. IEEE, Vol. 67, pp. 786-804, May 1979.