Author: |
陳彥儀 Yen-Yi Chen |
---|---|
Thesis Title: |
公車專用道之監視系統 A Restricted Bus-lane Monitoring System |
Advisor: |
陳世旺
Chen, Sei-Wang |
Degree: |
碩士 Master |
Department: |
資訊工程學系 Department of Computer Science and Information Engineering |
Thesis Publication Year: | 2010 |
Academic Year: | 98 |
Language: | 中文 |
Number of pages: | 71 |
Keywords (in Chinese): | local optimal thresholding 、深色區域 、亮部區域 、occlusion 、候選車輛 、deformable model 、攝影機校正 |
Keywords (in English): | local optimal thresholding, dark regions, bright regions, occlusion, vehicle candidates, deformable model, camera calibration |
Thesis Type: | Academic thesis/ dissertation |
Reference times: | Clicks: 66 Downloads: 2 |
Share: |
School Collection Retrieve National Library Collection Retrieve Error Report |
許多城市均有設置公車專用道,它使得公車線路更快且更可靠,然而若有未經授權的車輛駛入公車專用道,將使得公車專用道的好處無法完全實現。本系統主要的目的就是提供一個自動化的系統去偵測公車專用道上非法闖入的車輛,內容大致上分為兩部分:車輛偵測以及非公車辨識。首先做車輛偵測,我們利用local optimal thresholding的方法找出兩種可靠的區域特徵,分別為dark和bright regions,由於區域特徵為局部性特徵,較不受occlusion的影響,又每一輛車都有dark和bright regions,我們可以藉由合併dark和bright regions來偵測車輛。之後對於每個前景物判斷其是否有occlusion的情形,若有則進行切割。當所有前景物都進行切割或不切割的處理之後,得到的每一個object即視為一台車輛。接著在非公車偵測中,對每一台車輛套用一個deformable model,依據在camera calibration中得到的攝影機參數,換算出每一車輛在真實世界的長、寬、高,之後便可辨認出公車專用道上的車輛是否為公車,若為非公車,則須將之標示出來。
This paper presents a system for monitoring a bus lane to ensure that it remains effective for improving mass transportation time. The system detects unauthorized vehicles driving in the bus lane. Processing is done by two major steps: vehicle location and non-bus detection. In vehicle location, we search for dark and bright regions in the input image. Dark and bright regions are next paired if they comply with predefined criteria. Vehicle candidates are hen formed from paired dark-bright regions. In non-bus detection, for each vehicle candidate we first examine whether or not occlusions occur within the candidate. Region segmentation will be carried out if occlusions are detected. Thereafter, for each individual region we fit with a generalized deformable model. Bus and non-bus taxonomy is accomplished primarily on the basis of the size of the resultant model, which is a function of image row determined during camera calibration.
[Cha93] Charkari, N.M., Mori, H., “A new approach for real time moving vehicle detection,” Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems, Vol. 1, pp. 273 – 278, 1993.
[Fun03] Fung George S. K., Yung Nelson H. C., Pang Grantham K. H., “Camera calibration from road lane markings,” Opt. Eng. SPIE, vol. 42, no. 10, pp. 2976–2977, Oct. 2003.
[Gup02] Gupte, S., Masoud, O., Martin, R.F.K., Papanikolopoulos, N.P., “Detection and classification of vehicles ,“ IEEE Transactions on Intelligent Transportation Systems, Vol. 3, pp. 37 – 47, 2002.
[Lai01] Lai, A.H.S., Fung, G.S.K., Yung, N.H.C., “Vehicle type classification from visual-based dimension estimation,” Proceedings of Intelligent Transportation Systems, pp. 201 – 206, 2001.
[Lee06] Lee Deaho, Park Youngtae, “Robust vehicle detection based on shadow classification,” IEEE Conference on International Pattern Recognition , Vol. 3, pp. 1167 – 1170, 2006.
[Lin09] Lin Shanming, Tang Jun, Zhang Xuewu, Lv Yanyun, “Research on traffic moving object detection, tracking and track-generating” IEEE International Conference on Automation and Logistics, pp. 783 – 788, 2009.
[Mor94] Mori, H., Charkari, N.M., Matsushita, T., ”On-line vehicle and pedestrian detections based on sign pattern,” IEEE Transactions on Industrial Electronics, Vol. 41,pp. 384 – 391, 1994.
[Mos07] Mosabbeb, E.A., Sadeghi, M., Fathy, M., Bahekmat, M., “A low-cost strong shadow-based segmentation approach for vehicle tracking in congested traffic scenes,” International Conference on Machine Vision, pp. 147 – 152, 2007.
[Pal93] Nikhil R. Pal, Sankar K. Pal, “A review on image segmentation techniques,” Pattern Recognition Society, Vol. 26, pp. 1277– 1294, 1993.
[Pan04] Pang, C.C.C., Lam, W.W.L.; Yung, N.H.C., “A novel method for resolving vehicle occlusion in a monocular traffic-image sequence,” IEEE Transactions on Intelligent Transportation Systems, Vol. 5, pp. 129 – 141, 2004.
[Pan07-1] Pang, C.C.C., Tan Zhigang, Yung, N.H.C., “A methodology for resolving severely occluded vehicles based on component-based multi-resolution relational graph matching,” International Conference on Machine Vision, pp. 141 – 146, 2007.
[Pan07-2] Pang, C.C.C., Lam, W.W.L., Yung, N.H.C., “A Method for Vehicle Count in the Presence of Multiple-Vehicle Occlusions in Traffic Images,” IEEE Transactions on Intelligent Transportation Systems, Vol. 8, pp. 441 – 459,2007.
[Par01] Park Youngtae, “Shape-resolving local thresholding for object detection,” Journal, Pattern Recognition Letters , 22(8), south korea, PP. 883–890, 2001.
[Sad06] Sadeghi, M., Fathy, M.,”A Low-Cost Occlusion Handling Using a Novel Feature in Congested Traffic Images,” IEEE Conference on Intelligent Transportation Systems, pp. 522 – 527, 2006.
[Wu08] Wu Wei Zhang, Q.M.J., Xiaokang Yang, Xiangzhong Fang, “Multilevel Framework to Detect and Handle Vehicle Occlusion,” IEEE Transactions on Intelligent Transportation Systems, Vol. 9, pp. 161 – 174, 2008.