An algorithm for Displacement Extraction based on GPS and Monocular Vision Images
Joonhoo Lim, Hee Sung Kim, Je Young Lee, Kwang Ho Choi, Hyung Keun Lee
In the case of urban environments, GNSS(Global Navigation Satellite System) has drawbacks. Most of all, GNSS cannot provide positioning solution continuously due to insufficient number of visible satellites. By adopting hybrid positioning methods it is possible to overcome shortcoming of GNSS and improve positioning accuracy. A typical configuration for this purpose is the combination of GNSS with vision sensors. As an extension of existing hybrid positioning methods, this paper proposes an efficient method that combines a GNSS receiver with a monocular vision sensor. The proposed method can provide absolute position solutions continuously even if GNSS cannot provide positioning solutions intermittently. To evaluate the performance of the proposed method, an experiment result with field-collected real measurements is demonstrated.
Keywords: hybrid positioning method, GNSS, vision sensor
|