跳到主要內容區塊

 

104 影像辨識技術應用於室內導航系統之研究

出版年度:104

 

論文名稱:影像辨識技術應用於室內導航系統之研究

 

研究生:林善智   指導教授:游耿能

 


論文簡介:

  隨著行動技術的快速進步,在過去十年期間關於地理導航在行動裝置上的應用程式來幫助人們導航已經有了廣泛的探索。然而,這些應用程式大部分都著重在輔助室外導航。本研究提出了創新的辦法來輔助室內導航的部份。我們探索各式各樣的方法來幫助行動裝置上的室內導航。然後我們構建了一個原型來輔助室內導航。在我們的原型裡,使用名為尺度不變特徵變換 (SIFT) 的影像識別技術來進行室內定位和 A * 演算法用於尋找路徑。這個原型可以簡單的被安裝在智慧型手機上來輔助室內導航。我們在新的會議上做實地測試,讓第一次來參與的人在舉行會議的大樓樓層進行實測。大部份的人在自己的手機上都試圖使用這個系統並且給予寶貴的回饋。普遍認為使用本系統是被接受和有趣的。基於這些回饋和使用者的建議,原型的一些改進已經被實現,但是可用性和評價需要進一步探討。 With the rapid progress of mobile technology, applications of mobile devices about geographic navigation support to help people move around have been extensively explored during the last decade. However, most of these applications are focused on supporting outdoor navigation. This study presents a novel way to support indoor navigation. We explore various approaches to help the indoor navigation with mobile devices. We then construct a prototype system to support indoor navigation. In our prototype, the image recognition technology named Scale Invariant Feature Transform (SIFT) is employed to locate the indoor position and the A* Algorithm is utilized for pathfinding. The prototype can be easily installed on smartphones for supporting indoor navigation. We did a field test on a real occasion of guiding new meeting attenders around the designated floor of a building. Most users with relevant devices had tried to use the system and delivered precious feedbacks. The general responses about the usage of the system are acceptable and interesting. Based on these feedbacks and users’ suggestions, some improvements of the prototype has been implemented, but the usability and evaluation need to be further explored.

瀏覽數