[Back]

G. Retscher, W. Leitner:
"Kinematic Trajectory Determination of MMS (Mobile Mapping Systems): Results of Simulation Studies";
Poster: Symposium on Geodesy for Geotechnical and Structural Engineering, Berlin; 2002-05-21 - 2002-05-24; in: "Geodesy for Geotechnical and Structural Engineering II", Berlin (2002), ISBN: 3-9501492-1-x; 535 - 538.

The demand for accurate, economic, fast and semi-automatic or automatic collection of spatial data, can be seen as the main motivation for the development of Mobile Mapping Systems (MMS). Especially MMS on vehicles have been developed for fast data acquisition of road features and roadside objects on the few decimetre to meter accuracy level for road databases or GIS. There are quite a number of systems worldwide available that are either in the stage of development or already commercially employed.

For continuous vehicle position determination in most systems a combination of satellite assisted navigation using GPS and inertial navigation systems (INS) is employed. This GPS/INS combination is supported by additional sensors, such as odometer or speedometer, barometer, compass, inclinometer and temperature sensors, to increase positioning accuracy and reliability. For data acquisition usually CCD cameras, S-VHS cameras, laser scanners, etc. are employed. The selection of sensors depends mainly on the system requirements, such as accuracy, reliability, operational flexibility and range of application. The sensors are usually mounted on a common platform and by synchronizing their data streams accurately, the solution of a specific problem is possible by using data from one integrated measurement process only.

Several transformations of the various sensor observations have to be performed to match the data in a common coordinate frame. As the GPS data is represented in the WGS-84 coordinate frame, this system is used mostly as the common coordinate frame for all observation data. The trajectory of the vehicle has to be estimated in real-time from all observations of the employed navigation sensors. Therefore a Kalman filter approach is mostly used, since this filter is particularly suited for on-line evaluations. In the diploma thesis of Leitner conducted on the Department of Applied and Engineering Geodesy of the Vienna University of Techology, Austria, an evaluation software package has been developed for optimal estimation of the vehicle’s trajectory for a MMS using observations of GPS, INS, dead reckoning sensors (e.g. odometer and digital compass) and barometer. The performance of the software package has been tested based on simulation studies. Test calculations were performed for three different types of MMS that achieve different levels of positional accuracies for the surveyed object points, i.e., a system with high positional accuracy where the standard deviation for the object points is less than 0.3 m, up to a maximum of 1 m for a system in the medium range and less than 1 m for a low accuracy system respectively. To simulate real world conditions during the survey, measurement runs with constant vehicle speed are compared to runs with changes in vehicle acceleration (acceleration and braking) as well as sensor failures are simulated, such as the loss of lock for GPS positioning due to obstruction of the GPS signals. Examples of the main results of the calculations are presented in the paper.