Observability analysis of rotation estimation by fusing inertial and line-based visual information: A revisit

This note presents a simple approach to the observability analysis of the rotation estimation using line-based dynamic vision and inertial sensors. The problem was originally raised and formulated in Rehbinder, and Ghosh [2003. Pose estimation using line-based dynamic vision and inertial sensors. IEEE Transactions on Automatic Control, 48(2), 186–199.] where the unobservable subgroup was derived using complex matrix manipulations. By solving linear quaternion equations and using set operations, we not only successfully obtain the same result but also naturally extend it for the case with linearly dependent lines. The development in this note is more straightforward and gives rise to a clearer picture of the problem.