Localization of mobile robots with omnidirectional vision using Particle Filter and iterative SIFT

Scale Invariant Feature Transform on the Sphere: Theory and Applications

“…This kind of sensors has a lot of applications, such as video surveillance (Boult et al 2001) or object tracking (Chen et al 2008), and their use has become very common in robot navigation (Menegatti et al 2006) and in autonomous vehicles (Ehlgen et al 2008;Scaramuzza and Siegwart 2008). Interest points and local descriptors-based techniques, such as SIFT, have been applied to omnidirectional images due to their good performance in planar images (Goedeme et al 2005;Tamimi et al 2006;Valgren and Lilienthal 2007;Scaramuzza and Siegwart 2008). Recently, several efforts have been made to develop algorithms specifically designed to treat these omnidirectional images (Bogdanova et al 2007;Hadj-Abdelkader et al 2008).…”

“…in cylindrical coordinates) and then to apply the conventional SIFT algorithm. In fact, using the same reasoning, the classical SIFT has been applied to unwrapped omnidirectional images (Goedeme et al 2005;Tamimi et al 2006;Valgren and Lilienthal 2007;Scaramuzza and Siegwart 2008). The difficulties in this case come when there is information in the extremities of the omnidirectional image.…”

Monte Carlo Localization using 3D texture maps

“…SIFT-based MCL [3] is also compared to the proposed approach. The same motion model in Section III-A is adopted and the weight of each particle is computed as follows,…”

“…There are two generalized methods for a mobile robot to perform localization in a given map: metric localization [1]- [3] and topological localization [4], [5]. Given a 2D map, the result of metric localization is a position (x, y) and heading θ in the map.…”

“…with expected measurements. According to the given map, proximity sensors [2] or cameras [3], [8] are used to perceive the environment. A texture map is a metric map including environmental appearance.…”