View-based localization using panoramic images

Thanks to panoramic images, one can exploit image indexing techniques to qualitatively localize the rover as it re-traverses an already explored area. Our approach relies on the possibility to efficiently and robustly compute the resemblance between panoramic images, indexing them by histograms of local appearances. A database of image indexes is dynamically built during rover motions: when the rover re-perceives an already crossed area, it matches the current image with the stored ones (place recognition), and thus gets a qualitative estimate of its position.

The principle of building and use of the histograms is presented in the figures below [Gonzalez 2002a] [Gonzalez 2002]. The qualitative position is computed using the distance between the histograms that represents the image. The robot orientation is then computed by correlating thin rings extracted from the images.

Learning step: building of the database




Localization step