整合GPS/GIS於嵌入式平台之主動式車輛導航系統

Abstract

本論文提出一個整合無線網路與路口影像及交通壅塞資訊的主動式導航系統，回顧目前在市面上常見的車輛導航系統，在路徑規劃上通常以最短距離為考量而沒考慮實際交通狀況，因此所規劃的最短路徑亦有可能是最壅塞的路段；另外在車輛防盜監控功能方面，大多以傳送文字簡訊方式告知使用者，對於失竊的車輛並無法提供立即且有效的幫助。為了改善上述問題，本論文採用TI OMAP5912嵌入式平台建置一導航系統，整合GPS接收器、USB 攝影機、觸控式面板以及3.5G/3G無線模組等周邊，搭配全球定位系統與地理資訊系統，並使用地圖比對與地圖旋轉演算法來提升顯示流暢性與定位準確性，同時改良Dijkstra's路徑規劃演算法，加入路口壅塞資訊來規劃考量實際路況之最佳路徑，另外亦可透過3G無線網路將導航的電子地圖、車內影像傳送至使用者的行動裝置上，來提升車輛行動監控與防盜追蹤能力，在行駛過程中也提供觀看即時路口影像的功能，以讓駕駛者隨時隨地掌握交通的狀況。

In this thesis, an active vehicle navigation system integrating wireless networks, intersection images and traffic information is proposed. In general, the conventional navigation system plans the shortest path without considering the real traffic status of the road. That is to say, the planned route might be in the traffic jam. Moreover, the car surveillance system usually gives the alarm as soon as the car is stolen or the abnormal event is happened by sending warning messages to the driver, which can not provide an immediately assistance. In order to solve the problems addressed, the proposed system is build in the TI OMAP5912 platform equipped with a GPS receiver, a USB camera, a touch screen and a wireless module. By using the map matching and map rotation algorithm, the system provides accuracy positions of the car in the map and good quality of map display. For active path planning, Dijkstra's shortest path algorithm is enhanced to provide an optimum evaluation with the actually real-time jammed traffic information. Furthermore, to supply the driver with the mobile surveillance and burglarproof tracking, the navigation maps and images inside the car are delivered to the mobile device via 3G wireless networks anytime and anywhere. Finally, real-time streaming images in the intersection are transmitted via 3G wireless networks to the driver to realize the true traffic status.