This research explores robotic mapping algorithms with a focus on sonar mapping for indoor environments. It presents the major challenges in the problem of robotic mapping as well as solutions developed by many successful researchers in the realm. Subsequently, it proposes a new approach developed on the basis of the robust occupancy grid model. One of the major weaknesses of the standard occupancy grid model is caused by the independence assumption. The independence assumption works similarly to solving a puzzle without a framing context, which could be extremely difficult and prone to errors. The proposed model, however, attempts to relax the independence assumption by providing neighboring grid cell information and maintaining dependency information among these neighboring cells. Therefore, overall the results of this approach are convincingly better and more accurate than the results of the standard model.
展开▼
