Abstract: | Global Navigation Satellite System (GNSS) is widely used for a variety of mobile applications. However, GNSS performance in urban areas and indoor environments is problematic because of signal blocking, attenuation and multipath. In those circumstances, auxiliary devices are needed for successful navigation. A beacon system might be one solution to this obstacle as it requires scattered signal sources. On the other hand, vision-based navigation does not require any infrastructure system. Due to this major reason, it would be a reasonably proper system for indoor and urban areas. Vision-based system performance highly depends on the distribution of feature points that are the points of interest in surrounding images. This paper presents a calculation algorithm of the feature point’s distribution coefficient; the idea of this coefficient is adapted from the DOP (Dilution of Precision) of the GNSS area. Each feature point is assumed to be a signal source (satellite for GNSS), and the value is then calculated using the GNSS DOP. This value can be used for predicting estimation error like the DOP. The proposed method is demonstrated by analyzing the variance of the estimated position of the vision-based navigation system. Using the proposed method, the optimal direction of a vision sensor (i.e., a camera) can be determined in user specific environments. Furthermore, the estimation error can be reduced in operation by tilting the camera to the optimal direction of value maximization. This trend and its characteristics are analyzed and shown by numerical simulation. |
Published in: |
Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010) September 21 - 24, 2010 Oregon Convention Center, Portland, Oregon Portland, OR |
Pages: | 2285 - 2289 |
Cite this article: | Won, D.H., Chen, S., Sung, S., Lee, Y.J., "Error Analysis of Vision-Based Navigation System Using GPS DOP Concept," Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010), Portland, OR, September 2010, pp. 2285-2289. |
Full Paper: |
ION Members/Non-Members: 1 Download Credit
Sign In |