Credibilist simultaneous Localization And Mapping with a LIDAR

Abstract

From the early beginning, the Simultaneous Localization And Mapping (SLAM) problem has been approached using a probabilistic background. A new solution based on the Transferable Belief Model (TBM) framework is proposed in this article. It appears that this representation of knowledge affords numerous advantages over the classic probabilistic ones and leads to particularly good performances (an average of 3.2% translation drift and 0.0040deg/m rotation drift), especially when it comes to crowded environment. By introducing the basic concepts of a Credibilist SLAM, this article aims at proving that the use of this new theoretical context opens a lot of perspectives for the SLAM community.

Credibilist SLAM performances with different laser set-ups

“…The reader might then find a detailed description of the complete process in the precedent author's work [1].…”

“…The Credibilist SLAM (C-SLAM) concept have been introduced in [1]. It is inspired by a ML-SLAM solution used by Q. Baig et al [17] and J. Xie et al [10] and adapted to credibilistic occupancy grid.…”

“…The proposition made by the authors in [1] was to use both the properties of Transferable Belief Models (TBM) [15] and the ML-SLAM architecture [10] in order to quantify the belief in a landmark to be static or not. This quantification thus avoids the above hypothesis and so leads to more robust SLAM algorithms in crowded environment.…”

Simultaneous Localization and Mapping: A Survey of Current Trends in Autonomous Driving

“…Landmark uncertainties are represented by the occupancy probability of the cell which makes updates on map parts possible. During the update step, the classical approach is to maximize the similarity between the measurement and the map like in [59] in the Bayes formalism or in [60] in the TBM context. Not mentioned yet, the use of RADARs for SLAM has been demonstrated with filter-based approaches in [61][62] [63].…”

“…The idea is to take advantage of the map building process to directly detect and track obstacles by analyzing observations which are not coherent with the vehicle displacement. Alternatively, credibilist approaches [60], which represent ambiguous information, can indirectly deal with dynamic obstacles. They do not detect or track these obstacles but instead treat them as conflicts which allows the algorithm to affect a very low weight to these observations when estimating the vehicle displacement.…”