Geo-Tagged 3D Geometric Modeling of Urban Structures by Mitigating Reflected GPS Signals Using a Laser Range Sensor

In this paper, estimation of GPS position in urban scenario is improved using statistical approach and by mitigating reflected GPS signals using geometry of environment around measurement point for 3D modeling. Statistical approach works by considering satellite conditions, distribution of the position estimates followed by RANSAC and Hill Climbing approach. The reflected signals that are major cause of GPS position error in the urban scenario are identified using laser range data taken with laser scanner and are sacked from estimation of position calculation. A weighted non-linear least square solution is obtained for the position estimation using trilateration excluding non line of sight satellites. Implausible results obtained from experiments done in urban scenario during a span of time and over a span of several weeks validate the applicability and the feasibility of proposed method.