應用混合車流路口影像擁擠指數的交通號誌轉換時間估算

Determination of traffic light durations using crowd index estimation from videos taken at street intersections with mixed traffics

張仲耀(Chung-Yao Chang)

指導教授 ： 鄭士康

國立臺灣大學/電機資訊學院/電信工程學研究所/碩士(2008年)

https://doi.org/10.6342/NTU.2008.00529

摘要

現今交通系統常常使用車流量來估計交通號誌時間：用先前量好的道路最大可通過車流量，與單位時間需要通過的汽車流量，來決定路口的紅綠燈時間設定。目前已有非常多的軟體，可以輸入路口車輛參數，進而幫助估計出路口交通號誌的時間，但是均只限於汽車專用道路，還沒有應用於混合車流道路的軟體。本論文主要是應用紅燈轉綠燈時，車流擁擠程度的路口影像，利用簡單的影像處理技術得到擁擠指數，再將路口所偵測到的擁擠指數(crowd index)對應迴歸曲線(regressive curve)找出對應時間，估計本次綠燈時間的變動，以改善路口交通的等待時間。由於不必直接計算車輛數，所以可應用於混合車流的路口。也因為可以動態調整，可以避免等紅燈的用路者，即使綠燈車道已沒車流，仍須繼續空等的情形。

關鍵字

數位影像處理 ； 混合車流 ； 擁擠指數 ； 交通號誌 ； 路口影像

並列摘要

Duration of traffic lights nowadays are usually estimated by traffic flow：Measure the maximum traffic flow which the intersection can handle, and calculate the traffic flow during different hours and decide duration of the traffic lights. There is plenty of software to help engineers to decide the duration of traffic lights on cars-only road as long as we know the number of cars, but there is not any software to help deciding duration of traffic lights on mixed traffics. The main idea of this thesis is using the image before traffic lights turning green to calculate crowd index via simple image processing. We use crowd index to get a regressive curve, and from the curve to find the changes of duration of green light. These changes can improve the waiting time. Since we do not need to calculate the number of cars, we can apply this method to mixed traffics. Because the final calculation can be done very quickly we can prevent the unnecessary waiting, when apply the system to real-life traffic lights.

並列關鍵字

digital image processing ； mixed traffics ； crowd index ； traffic lights ； videos taken at intersection

參考文獻

[1] R. Awan, and F. Rani, "Video Based Effective Density Measurement for Wireless Traffic Control Application." Emerging Technologies, 2007. ICET 2007. International Conference on, pp. 99-101, 2007.

[3] Y. C. Chen, and T. H. Chang, “A study on traffic parameter extraction from image deterctor at intersection ” National Taiwan University Master Thesis, 2008.

[5] C. C. Cheng, and T. H. Chang, “A Study on Traffic Parameters Extraction from Monitroing System via Image Processing,” National Taiwan University Master Thesis, 2006.

[9] R. Inigo, “Traffic monitoring and control using machine vision: a survey,” IEEE Transactions on Industrial Electronics, pp. 177-185, 1985.

[10] R. Cucchiara, M. Piccardi, and P. Mello, “Image analysis and rule-based reasoning for a traffic monitoringsystem,” IEEE Transactions on Intelligent Transportation Systems, vol. 1, no. 2, pp. 119-130, 2000.