我們參考了以影像做定位的一些研究，並在近幾年智慧型手機普及的基礎上，實作了一套以Android手機為運行環境，透過手機內建的相機進行即時影像分析，並融合手機中感測元件做姿態分析，完成實際可運行的定位系統，此系統的特色是不需要使用任何輔助裝置，可單獨在手機運行，環境建制費用極低，且可非常容易部署至一般消費者的智慧型手機上，實驗結果證明我們的定位系統在小範圍以及中範圍應用中提供優於GPS的定位精確度；在大範圍應用中，我們的系統也能夠提供足夠的精確度做為適地性服務及擴增實境領域的定位 。

We implemented a visual-based positioning system on the Android-based smartphone. This system uses only build-in camera and sensors in an Android phone. The system works with no other assisted equipments, so it is low-cost and very easy to deploy to consumers with a mobile device. In addition, we conducted small, median, and large-scale field experiments to examine the accuracy and stability of our system. The results indicate that our visual-based positioning system is able to provide a much better accuracy in small and median-scale experiments compared with the global positioning system. In large-scale experiments, the system is able to provide enough accuracy for applications such as virtual augmentations and location-based services.

[1] Anja Bekkelien, Michel Deriaz, and St ́ephane Marchand-Maillet. Bluetooth indoor positioning. 2012.
[2] David Bernstein and Alain Kornhauser. An introduction to map matching for per- sonal navigation assistants. 1998.
[3] Nick Bilton. Behind the google goggles, virtual reality. Published: February, 22, 2012.
[4] David M Chen, Georges Baatz, K Koser, Sam S Tsai, Ramakrishna Vedantham, Timo Pylvanainen, Kimmo Roimela, Xin Chen, Jeff Bach, Marc Pollefeys, et al. City- scale landmark identification on mobile devices. In Computer Vision and Pattern Recognition (CVPR), 2011 IEEE Conference on, pages 737–744. IEEE, 2011.
[5] Shane Colton and FRC Mentor. The balance filter. Presentation, Massachusetts Institute of Technology, 2007.
[6] Android Developers. Android developer tools (adt) plugin for eclipse. URl: http://developer.android.com/tools/, 2013.
[7] Android Developers. Sensors overview — android developers. http://developer. android.com/guide/topics/sensors/sensors_overview.html, 2013.
[8] Brian Ferris, Dieter Fox, and Neil D Lawrence. Wifi-slam using gaussian process latent variable models. In IJCAI, volume 7, pages 2480–2485, 2007.
[9] Aaron Ganick and Daniel Ryan. Light positioning system using digital pulse recog- nition, January 31 2013. US Patent 20,130,028,475.
[10] Paul C. Glasser. An introduction to the use of complementary an in- troduction to the use of complementary filters for fusion of sensor data. http://www.glassercommunications.com/paul/samples.htm.
[11] Harlan Hile and Gaetano Borriello. Positioning and orientation in indoor environ- ments using camera phones. Computer Graphics and Applications, IEEE, 28(4):32– 39, 2008.
[12] Yunye Jin, Hong-Song Toh, Wee-Seng Soh, and Wai-Choong Wong. A robust dead- reckoning pedestrian tracking system with low cost sensors. In Pervasive Computing and Communications (PerCom), 2011 IEEE International Conference on, pages 222– 230. IEEE, 2011.
[13] H Kawaji, K Hatada, T Yamasaki, and K Aizawa. An image-based indoor positioning for digital museum applications. In Virtual Systems and Multimedia (VSMM), 2010 16th International Conference on, pages 105–111. IEEE, 2010.
[14] Alasdair Mackintosh, Alexander Martin, Brian Brown, Christian Brunschen, and S Daniel. Zxing, open source library to read 1d/2d barcodes, 2012.
[15] Alessandro Mulloni, Daniel Wagner, Istvan Barakonyi, and Dieter Schmalstieg. In- door positioning and navigation with camera phones. Pervasive Computing, IEEE, 8(2):22–31, 2009.
[16] J Angelo Racoma. Indooratlas uses magnetic fields for location-awareness in buildings and large structures, 2013.
[17] wikipedia. Angle of view.
[18] wikipedia. Orthographic projection.
[19] wikipedia. Simple moving average.
[20] wikipedia. Spherical coordinate system.
[21] wikipedia. Taipei 101.