在今日科技文明社會中每天均需使用道路，所以普遍裝設了交通監視系統以便能取得即時的路況資訊，甚至偵測出意外之交通事故；針對此需求本論文提出了一個運用在市區交通監視上之即時車輛計數和分類演算法以滿足此方面之需要。
對於即時之市區道路視訊影像，本論文首先利用中間值濾波器選取合適亮度值建立一可定時更新之背景影像，並藉以擷取出前景移動物件經過相關前置處理後，根據移動物件區塊所在位置比對追蹤前張影像可能出現之範圍，使用其相互之距離、形狀及色彩資訊等特徵值移動物件作為車種區分的判斷和黏合狀況的偵測，以構設此一可區分車輛種類之追蹤演算法；同時對於前後車距過近、車輛暫停路中或機車行駛快車道以及蛇行等異常行駛狀況作分析以便監視可能發生之交通事故之狀況。由於此演算法是利用影像中移動物件可能出現範圍作預測，所以可減少不需要比對之移動物件之運算，因此甚為快速。
　　本論文所提出的車輛分類及計數演算法經不同路況之交通視訊影像實驗驗證，確實可運用於一般市區交通的各種路況，具有一定水準之準確率，除了可即時計算各種車輛數目，且對於異常駕駛狀況亦具追蹤之效果。

In nowadays, technology is civilized. Most people make use of road in their daily life every day, therefore; it’s simple to set traffic monitoring system for getting real-time traffic information even detecting traffic accidents.For the requirement, this paper proposed a Real-time Vehicle Classification and Counting Algorithm for Downtown Traffic Monitoring.
This way could reduce moving object’s tracking and matching time, when the traffic jam happened is more remarkable. First, utilize an median filter to establish a reliable background model , then extract the foreground object for adjacent downtown traffic image, after take out the foreground object, noise filtering and shadow removal process, according to the Boundingbox of pixel distribute information as a vehicle-kind judged and occlusion detect.Also, established the centroid of moving object to track and match, based on the location of object region to predict the multi segmented scene where we search, and proceeding compare the dual of boundingbox’s c distance and the average value error to decide whether both of them are the same moving object then finally statistic the quantity information of renewal car.
The paper what we proposed to the vehicle counting could apply on many conditions in the downtown traffic scene. The method keeps up good performance and could calculate car quantity fast. Besides counting vehicles number, the system also effective on detecting and tracking abnormal driving.

[1] Q.Zhou and J.K. Aggarwal, “Tracking and Classifying Moving Objects from Video,”Proceedings of IEEE Int.Workshop on PETS 2001,p44-p51

[2] Manchun LEI, Damien LEFLOCH, Pierre GOUTON and Kadder
MADANI, “A video-based real-time vehicle counting system using adaptive background method,” 2008 IEEE International Conference on Signal Image Technology and Internet Based Systems,P523-528

[3] Liu Bo and Zhou Heqin, “Using object classification to improve urban traffic monitoring system,” IEEE Int. Conf. Neural Networks 8 Si nal Processing Nanjing, China, December 14-17. 2003

[4] Thou-Ho Chen, Yu-Feng Lin, and Tsong-Yi Chen ,“Intelligent Vehicle Counting Method Based on Blob Analysis in Traffic Surveillanc,” IEEE2007 Innovative Computing,Information and Control,2007.ICICIC’07 Second International Conference p238-p241

[5] Shunsuke Kamijo, Yasuyuki Matsushita, Katsushi Ikeuchi, “Traffic Monitoring and Accident Detection at Intersections, ” IEEE2000 Intelligent Transportation Systems,IEEE Transaction,P108-118

[6] B.K.P. Horn and B.G. Schunck., “Determining optical flow, ” AI Memo 572. Massachusetts Institue of Technology, 1980,p78-84

[7] Cathey,F.W.;Dailey,D.j, “A novel technique to dynamically measure vehicle speed using uncalibrated roadway cameras, ”IEEE2005 Intelligent Vehicles Symposium,2005.IEEE,p777-p782

[8] D.Beymer and K.Konolige, “ A real-time computer vision system for measuring traffic parameters,”Proc.IEEE 1997, IEEE Computer Society Conference ,p495-p501

[9]C.Stauffer, W.E.L.Grimson,“Adaptive Background Mixture Models for Real-time Tracking,”IEEE 1999, Computer Vision and Pattern Recognition,1999.IEEE Computer Society Conference,p83-p88

[10] R.Mech and M.Wollborn, “A noise robust method for 2D shape estimation of moving objects in video sequences considering a moving camera, ”IEEE2003,Signal Processing, vol.66,p203-p217,Apr.1998

[11] Kunfeng Wang, Zhenjiang Li, Qingming Yao, Wuling Huang, and Fei-Yue Wang, “An Automated Vehicle Counting System for Traffic Surveillance,” IEEE2007,Vehicular Electronics and Safety, 2007,ICVES.IEEE International Conference,p1-p6

[12] 謬紹剛 譯, “數位影像處理-運用Matlab” 東華出版社, 2005

[13] 林傳生, “Matlab之使用與應用” 儒林出版社, 2004