Fusing Low-Cost Image and Inertial Sensors for Passive Navigation

Publisher Website
Google Scholar
Abstract
Navigation parameters (position, velocity, and attitude) can be estimated using optical measurements combined with an inertial navigation system. This can be accomplished by tracking stationary optical features in multiple images and using the resulting geometry to estimate and remove inertial errors. A critical factor governing the performance of image-aided inertial navigation systems is the robustness of the feature tracking algorithm. Previous research has shown the benefit of coupling the sensors at the measurement level using a tactical-grade inertial sensor. While the tactical-grade sensor is a reasonable choice for larger platforms, the greater size and cost of the sensor limits its use in smaller platforms. In this paper, an image-aided inertial navigation algorithm is implemented using a multidimensional stochastic feature tracker and low-cost sensors. The performance of the resulting navigation system is evaluated and compared. The fused image-inertial sensor is shown to outperform a free-running tactical-grade inertial sensor.

This publication has 4 references indexed in Scilit: