Augmented state Kalman filtering for AUV navigation

Addresses the problem of estimating the motion of an autonomous underwater vehicle (AUV), while it constructs a visual map ("mosaic" image) of the ocean floor. The vehicle is equipped with a down-looking camera which is used to compute its motion with respect to the seafloor. As the mosaic increases in size, a systematic bias is introduced in the alignment of the images which form the mosaic. Therefore, this accumulative error produces a drift in the estimation of the position of the vehicle. When the arbitrary trajectory of the AUV crosses over itself, it is possible to reduce this propagation of image alignment errors within the mosaic. A Kalman filter with augmented state is proposed to optimally estimate both the visual map and the vehicle position.

1 Introduction

Creating visual maps of the ocean floor is an important tool for underwater navigation. Consider an AUV equipped with a down-looking camera, which provides images of the seabed as the vehicle moves. The alignment of these images provides the necessary information to estimate the position and orientation of the vehicle [1], [2]. At the same time, warping the aligned images creates a visual map, known as mosaic, which can be used for planning future missions [3]. While the mosaic is constructed, the vehicle can localize itself in this map, following the Concurrent Mapping and Localization methodology [4], [5].