隨著GPS技術的普及，舉凡汽車導航裝置、智慧型手機、穿戴式裝置以及地理量測系統等需要定位的應用，都可以見到GPS定位技術的廣泛應用，其中以汽車導航裝置最為一般大眾所熟悉並獲得廣泛的商業成功。而汽車導航應用的場合主要在複雜的都市化環境，林立的高樓無可避免的會遮蔽GPS衛星訊號或造成多路徑反射，導致GPS接收機的定位誤差。

假設從GPS接收機可得到衛星資料(仰角、方位角、訊雜比)及經緯度座標表示目前使用者的位置，並且能夠得知現在GPS接收機受到的多路徑效應影響程度，就可以去推論在定位的經緯度解上造成多大的誤差。

本論文利用安裝在車頂的360度視角魚眼攝影機往天空拍下影像，使用二值化影像辨識可直視天空的區域，並配合GPS接收機收到的NMEA資料，可得知當時的衛星分佈、被建築物遮蔽的衛星，以及道路兩旁建築物造成的多路徑效應環境為何? 可以初步判斷出GPS接收器目前的位置是有誤差的，進而推估這樣環境下的誤差，再利用REVERSE DGPS 演算法修正虛擬距離的誤差量後，可以算出位置偏離的方向，進而修正誤差距離而得到較正確的位置。

本論文希望在汽車導航時，因為多路徑效應而偏離正確路徑時，能夠適度把它修正回正確路徑，經由實驗的驗證，的確可以降低誤差。

As GPS technology becomes more popular, its application becomes widely adapted in products across the market. GPS technology can be found in products such as automobile navigation systems, smart phones, wearable devices, and geographical surveying equipment. Among these, automobile navigation systems are the most widely known, with widespread commercial successes. As automobile navigation systems are primarily used in highly urbanized areas, high-rises will inevitably disrupt GPS signal acquisitions due to obstruction of satellites or will cause multipath effect phenomenon when signals are reflected from the surfaces of surrounding buildings. Hence, inaccurate positioning will be generated by GPS receivers. We suppose GPS receiver get satellites information(Elevation、Azimuth、C/No) and a coordinate of latitude & longitude to indicate user’s current position, along with current level of multipath effect, therefore we can make an inference the multipath pseudorange error from positioning calculation.
This study used a fisheye camera mounted atop a vehicle to capture 360° aerial views of the sky via the binary image thresholding technique to identify boundaries of the sky with direct line-of-sight coverage. By incorporating NMEA data, the GPS receiver acquired the satellite aerial position, identified those that were obscured, and evaluated how surrounding buildings were contributing to the multipath effect. Because current position errors in GPS receiver could be preliminarily determined,multipath pseudorange error caused by surrounding interference effect was estimated. REVERSE DGPS algorithm was then applied to correct errors in pseudorange differences. Next, position deviations in direction and distance were calculated to obtain a more accurate position.
This study aims to provide a method to help a vehicle adhere to its proper course when navigating through areas with obstructed satellite signals and multipath effect that may potentially deviate the vehicle from its desired route. The experimental results show that GPS positioning errors can indeed be reduced using this technique.

[1] 莊智清，”衛星導航”，全華科技圖書，台北，2012
[2] Tsui, J. B. Y., "Fundamentals of global positioning system receivers, " Wiley-Interscience, 2006
[3] Grewal, M. S., Weill, L. R., & Andrews, A. P., ”Global positioning systems, inertial navigation, and integration”, John Wiley & Sons., 2007
[4] Groves, P. D., Wang, L., & Ziebart, M., ”Shadow matching: Improved GNSS accuracy in urban canyons,” GPS World, 23(2), 14-18., 2012
[5] Iwase, T., Suzuki, N., & Watanabe, Y., ”Estimation of Multipath Range Error for Detection of Erroneous Satellites”, In Proceedings of the 23rd International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS 2010) (pp. 2228-2235)., Oct 2001
[6] Peres, T. R., ”Multipath Mitigation Techniques Suitable for Low Cost GNSS Receivers [D],” Instituto Superior Técnico, 20-56., 2008
[7] 李原綜(1995)，”即時性DGPS載具定位系統”，碩士論文，國立台灣科技大學機械工程研究所，台北，1995
[8] 劉坤賦，”GPS信號品質量測與分析，碩士論文，國立台灣科技大學機械工程研究所，台北，2007
[9] u-blox document, “Datum Transformations of GPS Positions,” http://www.u-blox.ch.Application Note. 5th, July 1999
[10] National Marine Electronics Association. ,“NMEA 0183--Standard for Interfacing Marine Electronic Devices. NMEA., “, 2002
[11] 洪維恩，”MATLAB7程式設計入門篇”，旗標出版股份有限公司，2012
[12] Otsu, N., “A threshold selection method from gray-level histograms.,”Automatica, 11(285-296), 23-27., 1975
[13] 陳建隆，”應用改良式經驗模態分解法於消除文件影像中的不良光照現象”，碩士論文，國立中央大學資訊工程學系研究所，桃園，2009
[14] Mohammed, N. Z., Ghazi, A., & Mustafa, H. E., ”Positional accuracy testing of Google Earth,” International Journal of Multidisciplinary Sciences and Engineering, 4(6). , 2013
[15] Nelson, W., ”Use of circular error probability in target detection (No. MTR-10293)”, MITRE CORP BEDFORD MA., 1988
[16] Veness, C., “Calculate distance, bearing and more between Latitude/Longitude points,” http://www. movable-type. co. uk/scripts/latlong. html., 2010