PE&RS November 2001

Direct Georeferencing

The inertial navigation system (INS) was first demonstrated in 1949 by C.S. Draper, and since then has been the preferred method of navigation on military and commercial aircraft whose mission profile excludes dependency on external references or signal sources. In addition, the INS provides a complete navigation solution that includes position, velocity, attitude (roll, pitch, heading), accelerations, and angular rates of the vehicle or platform being navigated. During the 1970s, the gimbaled INS gave way to the strapdown INS with the advent of the ring laser gyro and a sufficiently powerful avionics computer that could perform the numerically intensive high rate strapdown navigation computations.

Aided INS (AINS) refers to the estimation of the INS navigation and inertial sensor errors and correction of the navigation solution through one of two methods. The first method corrects the output navigation solution of the free-inertial INS, and is called “feedforward” or “output error control.” The second method corrects the integration processes in the INS navigation computation, and is called “feedback” or “closed-loop” regulation of INS errors. The latter method can be used to implement mobile alignment of the INS. The INS errors are typically estimated by a Kalman filter. This technology was first developed for, and deployed in, military aircraft, beginning in 1968. More recently, the U.S. Air Force Embedded GPS-Inertial (EGI) program has spawned a standard GPS-AINS configuration for military navigation applications.

The geomatics community first considered the INS or AINS as a survey instrument during the 1970s. The INS’s autonomous navigation capability allows it to provide position data in circumstances where traditional survey methods are cumbersome, slow, or not feasible. Examples are land vehicle survey and seismic survey. Before the advent of GPS, the AINS survey systems typically used zero velocity updates to control the position error drift of the INS. The INS attitude solution was used to measure the orientation or pointing angles of airborne remote sensors such as multi-spectral scanners. The drawbacks of the INS for survey applications were its size, weight, and cost. A typical ring-laser gyro (RLG) INS during the period 1975-1985 weighed 47 lbs, had a volume of 1420 inches3, and cost close to $180K.

The advent of GPS introduced new possibilities in AINS for survey applications. For the first time, GPS provided an easily accessible and continuous stream of position and velocity measurements whose errors were typically noisy but bounded in magnitude. These error characteristics complemented an INS’s smooth but unbounded error characteristics, and through proper Kalman filter design and overall system integration a GPS-aided INS could exhibit the best attributes of both systems, i.e. smooth and bounded error characteristics. Moreover, GPS-aided INS provided continuous alignment while GPS was available so as to achieve highly accurate orientation data, and free-inertial “coasting” through GPS outages so as to achieve continuous position and orientation data.

GPS-aided INS allowed the implementation of the direct georeferencing concept for remote sensing applications. First, it has been experimented with independently by The University of Calgary and The Ohio State University in the late 1980s and early 1990s (for example, Schwarz, et al., 1993). Direct georeferencing (DG) is the direct measurement of the position and orientation of a mobile sensor, such as a camera, for the purpose of image stabilization, image motion compensation, multiple image collation and assignment of geographic coordinates to features in the image and possibly each image pixel. It was first put into practice in several experimental systems such as airborne multi-spectral scanners and laser altimeters. The orientation sensor typically was an aircraft INS that navigated free-inertially during the data acquisition. These systems demonstrated the DG concept, but were impractical because of the large size and high cost of the INS. The INS typically had to be a high accuracy instrument that used ring laser gyros and pendulous accelerometers to achieve the orientation accuracy needed for acceptable image stability and georeferencing. Such an INS was quite large and heavy and hence not easily mounted to smaller instruments. Furthermore, the RLG dither introduced short-term noise into the orientation parameters that compromised the photo-to-photo orientation accuracy.

A practical DG system became feasible with precise GPS positioning capability through integer ambiguity resolution and the arrival of strapdown inertial measurement units (IMU) for tactical and motion sensing applications in military weapons systems. A “tactical” IMU is designed for short-term missile guidance. It is required to be small and light in order to fit into an air-to-air missile case, and relatively inexpensive in order to be affordable in an expendable weapon. It uses low cost inertial sensors such as fiber-optic gyros (FOG) to meet a 1-10 degree/hour performance requirement. Such an IMU is impractical in a free-inertial navigation application. Its gyro errors would drive the position error drift to upwards of 100 nautical miles per hour. It is however suitable in a GPS-AINS, which calibrates the inertial sensor errors on an ongoing basis. The small size allows the IMU to be mounted close to a reference point being instrumented, for example inside the lens cone of a film camera near its perspective center such as that of the Z/I Imaging DMC camera and the LH-Systems ADS40 camera.

Applanix was the first organization to offer for sale a GPS-aided INS product specifically for commercial airborne surveying applications. The Position and Orientation System for Airborne Vehicles (POS/AV) for airborne remote sensing was released in 1996 (Hutton et al., 1997). Since then, airborne photogrammetrists have adopted the Applanix POS/AV to directly measure the exterior orientation parameters of each photo to either supplement or replace aerotriangulation. Such a capability has translated into significantly lower post-mission processing costs, which is the key value proposition of direct georeferencing. Other organizations have recognized the technology advance and the business potential, and have fielded competing products. DG systems for photogrammetry and remote sensing have moved from experimentation to mainstream usage. For example, the new generation of digital cameras from Leica and Zeiss rely on direct georeferencing to provide image stability and orthometric data for digital elevation modeling, and therefore include embedded Applanix POS/AV systems.

The focus of current development in GPS AINS for direct georeferencing is improved accuracy and improved ease of use. A GPS AINS for DG applications is critically dependent on the accuracy of GPS. Improved accuracy will come from improvements in IMU performance, better signal processing algorithms, and improved quality control. Improved ease of use will come from a new generation of GPS processing methodologies that include network receiver processing and the use of permanent GPS receiver arrays such as the International GPS Service for Geodynamics (IGS) and the U.S. National Geodetic Survey (NGS) Continuously Operating Reference System (CORS). For details, see Mostafa and Hutton (2001) and Bruton et al., 2001.

Further Reading
Hutton, J., Savina, T., and Lithopoulos, L., 1997. Photogrammetric applications of Applanix’s position and orientation system (POS), ASPRS/MAPPS Softcopy Conference, Arlington, Virginia, July 27 – 30 1997.

Mostafa, M. and J. Hutton, 2001. Airborne Kinematic Positioning and Attitude Determination Without Base Stations, Proceedings of International Symposium on Kinematic Systems in Geomatics and Navigation (KIS 2001), Banff, Canada, June 5-8.

Schwarz, K.P., M.A. Chapman, M.E. Cannon, and P. Gong, 1993. An integrated INS/GPS approach to the georeferencing of remotely sensed data, Photogrammetric Engineering & Remote Sensing, 59(11): 1167-1674.

Top

Home