Lattice independence and vision based mobile robot navigation

Strong Lattice Independence implies Affine Independence. Affine Independent sets of vectors define a convex polytope and if this polytope is a good approximation to the convex hull of a set data points, we can use them to represent the data points through their convex coordinates. This representation can be used as a feature extraction or dimensionality reduction method. Morphological Associative Memories (MAM) have been proposed for image denoising and pattern recognition. Recent works show that, by construction, Autoassociative Morphological Memories (AMM) are composed of lattice independent vectors. After a transformation these vectors can be shown to be a good approximation to the data convex hull, and therefore as a candidate set of points for convex coordinate representation of the data. In this paper we present some results on the task of visual landmark recognition for a mobile robot self-localization task improving previous results using AMM.