摘要：This paper presents an approach of improving navigational data in mobile mapping by using additional photogrammetric measurements. Images taken during a mobile mission are considered as support of GPS/INS while the consumer grade inertial MEMS unit used in the presented mobile system provides only poor estimation of position and attitude even with GPS aid. Nevertheless, this estimation provides initial values for image processing. Simultaneous bundle adjustment could be done easily with approximate values obtained by GPS/INS solution. The adaptation of our algorithm to meet the case of low cost land mobile mapping is also explained. The feedback to the trajectory determination is discussed from a Kalman decentralized filter point of view. Automatic tie point object coordinates approximation depending on terrestrial laser point clouds is also presented in order to reach a global automatic data processing. The precision of the solution is discussed and assessed by comparing the resulting point clouds with more accurate data sources.
