電腦視覺於機車相對距離之空間呈現

機車在東南亞地區不僅常見且為高使用率之運輸工具，因其高機動性及便利停放之特性，機車持有比在我國居高不下。然而，機車騎士暴露於車體外之比例高，且常在車流中鑽行以及騎士之車道觀念較薄弱的情況下，造成機車之安全程度較其他運具低。因此本研究希望透過電腦視覺，自動觀察特定路段或路口之機車微觀駕駛行為，並擷取巨觀車流之資料。在本研究中，攝影機架設於車道上方，經由機車之俯視畫面偵測路面中之機車並取得位置。方向梯度直方圖以及支持向量機被用來偵測錄影機所錄畫面中機車之位置，並應用卡曼濾波及匈牙利演算法，串聯機車之軌跡。透過本研究的自動化系統，取得車流之資訊，其中包括(1)在特定車道或路口之機車軌跡，(2) 機車之交通流量及平均速度，(3)縱向及橫向之機車相對空間累積位置圖之特徵分析，(4)相對速度快之機車及相對速度慢之機車與周邊機車相對空間位置圖之特徵分析。本研究透過真實之拍攝影片，驗證路段之機車安全指標，此研究之延伸，可能被運用於機車安全之路面幾何危險偵測。

關鍵字

電腦視覺；方向梯度直方圖；支持向量機；卡曼濾波；機車；交通安全

並列摘要

Individual motorcycle behavior depends on the surrounding environment such as traffic flow and the geometric design of the road. For their own safety, motorcyclists react to the distance and velocity of other vehicles. Especially during rush hours, the traffic condition changes rapidly. By setting up a camera at the road intersection to record traffic flow videos, the relative spatial position and the motorcycles’ direction angle are measured by analyzing the data. The motorcycle behavior at the intersection is observed and analyzed through computer vision and image processing methods. The Histogram of Gradients (HOG) descriptor is adapted for the detection of motorcycles utilizing a Support Vector Machine (SVM), and the Kalman Filter is employed for the tracking of the motorcycles’ trace. Through this approach, we observe the motorcycles’ traffic flow and traffic characteristics such as: (1) motorcycle trajectories on specific road section, (2) motorcycle traffic volume and average speed, (3) the pattern of accumulated relative spatial position of all the motorcycles from the video along the horizontal axis and vertical axis, and (4) the pattern of accumulated relative spatial positions of relatively fast and relatively slow motorcycles from the video. We have taken real world videos for the validation of the proposed approach. The results of this study could serve as a reference for traffic safety guidance for motorcyclists and could potentially be applied for detection of danger road geometries.

並列關鍵字

Computer Vision ； Histogram of Oriented Gradient ； Support Vector Machine ； Kalman Filter ； Motorcycle ； Traffic Safety

參考文獻

Wang, J. M., Chung, Y. C., Lin, S. C., Chang, S. L., Cherng, S., & Chen, S. W. (2004). Vision-based traffic measurement system. Pattern Recognition, 2004. ICPR 2004. Proceedings of the 17th International Conference on, 4.

Chintalacheruvu, N., & Muthukumar, V. (2012). Video based vehicle detection and its application in intelligent transportation systems. Journal of transportation technologies, 2(04), 305.

Chiverton, J. (2012). Helmet presence classification with motorcycle detection and tracking. Intelligent Transport Systems, IET, 6(3), 259-269.

Coifman, B., Beymer, D., McLauchlan, P., & Malik, J. (1998). A real-time computer vision system for vehicle tracking and traffic surveillance. Transportation Research Part C: Emerging Technologies, 6(4), 271-266.

Gazis, D. C., Herman, R., & Rothery, R. W. (1961). Nonlinear follow-the-leader models of traffic flow. Operations research, 9(4), 545-567.

國際替代計量

電腦視覺於機車相對距離之空間呈現

全文下載

主題瀏覽