標題: | 一個利用行動感測裝置的即時坑洞偵測方法 A Real-Time Pothole Detection Approach using Mobile Sensing Device |
作者: | 林淳皓 羅濟群 資訊管理研究所 |
關鍵字: | 智慧型運輸系統;坑洞偵測;行動感測;Intelligent Transportation System;Pothole Detection;Mobile Sensing |
公開日期: | 2011 |
摘要: | 近年來隨著經濟的快速發展,而為了解決日益嚴重的交通問題,許多國家寄望能夠透過資訊與通訊科技的協助,讓交通資源能夠更有效的運用,這正是智慧型運輸系統(Intelligent Transportation System, ITS)發展的目的。ITS發展目標之一是增進交通安全,然而在交通意外中絕大部分是道路上的坑洞造成用路人傷亡,因此如能建立一套即時的道路坑洞偵測系統並予以用路人坑洞資訊,將能大幅提升ITS運用在交通上的安全性。此外,近年來利用行動式手持裝置來偵測道路坑洞的方式已越來越受到重視,這種方法不需要花費龐大的金額來架設以及維護額外的偵測裝置,再加上現在的社會裡幾乎每個人都有手機,因此以行動式手持裝置來偵測道路坑洞是非常全面的。本研究提出一個以行動感測為基礎的即時坑洞偵測方法,透過此偵測方法分析內嵌於手機內的加速度感測器資料來進行坑洞判定。本研究另外也利用歐拉角(Euler Angles)進行手機加速度正規化的動作,改善蒐集加速度資料時需要固定手機姿態角的限制以及利用空間內插法改善全球定位系統(Global Positioning System, GPS)定位誤差。在實驗中以實際的道路坑洞進行路測,其實驗的結果顯示本研究提出的坑洞偵測方法高達90%的精確率。 In recent years, fast economic growth and rapid technology advance have led to significant impact of the quality of traditional transport system. Intelligent Transportation System (ITS), which aims to improve the transport system, has therefore become more and more popular. The one of the aims of ITS is to enhance traffic safety, the driver is injured mostly due to the potholes on the road in traffic accidents. If it is possible, establish of real-time pothole detection system and share the pothole information with drivers, that will be improve the traffic safety of ITS. Besides, use the mobile device to detect potholes is more popular in recent years. That is not spending lot of money to maintain additional devices, and almost everyone has a cell phone in current society, thus use mobile devices to detect potholes are very comprehensive. In this paper, we propose a novel pothole detection method based on the mobile sensing. We analyze the accelerometer data and adopt pothole detection algorithm to obtain the pothole information and use Euler angles to normalize the accelerometer data of mobile device. Finaly, we use the spatial interpolation method to improve the errors of GPS. The result of experiments also show that our pothole detection approach can reach 90% precision is better than existing pothole detection approach. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT079934502 http://hdl.handle.net/11536/50124 |
Appears in Collections: | Thesis |