Localization technique based on the voxel matching mrthod in a crowded environment — High speed voxel mathing method by integral images

This paper describes a localization technique for a mobile robot in an environment with many unknown obstacles, such as pedestrians. To realize robust localization against unknown obstacles by using a particle filter, a space observation model has conventionally been used. However, this approach requires a lot of computation time for matching voxel. Therefore, in this study, we propose a fast matching method using the integral image. In the proposed method, it is considered the spatial information defined by the voxel is made up of a plurality of grid map layer. By calculating the inconsistency of spatial information by the integral image for the grid map for each layer, and realize high-speed processing time.