協同式交通號誌偵測系統

Cooperative Traffic Light Detection System

李岱旻(Tai-Ming Li)

指導教授 ： 簡榮宏

國立交通大學/資訊學院/網路工程研究所/碩士(2014年)

https://doi.org/10.6842/NCTU.2014.01090

摘要

研究顯示車輛在行駛的過程中過多的停等，會額外增加油耗且排放更多的二氧化碳。若車輛在接近交岔路口前，駕駛能事先獲得前方交通號誌的時相規劃資訊，則可藉由調整車輛的行駛速度，更加平順地通過路口，減少路程中不必要的停等。有鑒於近年來在辨識交通號誌的相關影像技術越發成熟，以及雲端運算平台的快速發展，我們設計了一套協同式交通號誌偵測系統，此系統透過收集分析來自各車輛在通過路口時所上傳的號誌偵測結果與其行車路徑，伺服器可以估算各路口號誌當下確切的時相，提供號誌時相資訊給其他來車，運用在相關行車應用中。經模擬的結果顯示，我們設計的系統能在平均二十分鐘內同步各路口的號誌時相，且能有效地過濾百分之八十以上的車輛的誤判資訊，提供後方車輛各路口誤差一秒以內的號誌時相資訊。

關鍵字

Intelligent Transport System ； traffic light detection ； traffic light phase ； video sensor

並列摘要

Researches show that frequent stop-and-go of vehicles increases fuel consumption and CO2 emissions. If drivers are aware of the traffic light phase information in advance, they can adjust speed of cars so as to cross intersection smoothly with less stopping. By taking advantages of progressive traffic light recognition technique and the highly development of cloud computing in recent years, we propose a cooperative traffic light detection system. In the proposed system, each vehicle submits its path and the detected traffic light status when it passes an intersection. The backend server collects those information, estimates traffic light phase of each intersection and then shares them with other vehicles. Simulation results show that the proposed system can filter wrong detecting results above 80%. Furthermore, it can provide accurate phase information with mean absolute time error lower than 1 second after average 20 minutes data collecting time.

並列關鍵字

智慧型運輸系統 ； 交通號誌偵測 ； 交通號誌時相 ； 影像感測器

參考文獻

[2]. T. Tielert et al., “The Impact of Traffic-Light-to-Vehicle Communication on Fuel Consumption and Emissions,” Internet of Things 2010, Tokyo, Japan, Nov. 2010.

[5]. Jeffrey Miller, “Vehicle‐to‐Vehicle‐to‐Infrastructure (V2V2I) Intelligent Transportation System Architecture”, IEEE Intelligent Vehicles Symposium, pp. 715‐720, June 2008.

[7]. B. Asadi and A. Vahidi, “Predictive cruise control: Utilizing upcoming traffic signal information for improving fuel economy and reducing trip time,” IEEE Transactions on Control Systems Technology, 2010.

[8]. P. Schuricht, O. Michler, and B. Baker, “Efficiency-increasing driver assistance at signalized intersections using predictive traffic state estimation,” 14th International IEEE Conference on Intelligent Transportation Systems, pp. 347–352, 2011.

[9]. M. Kerper, C. Wewetzer, A. Sasse, and M. Mauve, “Learning traffic light phase schedules from velocity profiles in the cloud,” in New Technologies, Mobility and Security (NTMS), May 2012, pp. 1–5.2