Pre-Mapping System with Single Laser Sensor Based on Gmapping Algorithm

Abstract

Knowing physical structure of the unknown environment is a requirement in some natural disaster situations such as earthquake, fire accidents or flood. Many existing algorithms in literature have been developed for the purpose of extracting map of unknown environment. These algorithms called as SLAM (Simultaneous Localization and Mapping). But still there are problems in SLAM to be solved. One of these problems referred in literature as chicken-egg problem. Another struggling problem is requirement of more parameters to execute and generate accurate map and pose estimation. That makes systems' structures more complicated and hierarchical. In this study, parameters of Gmapping SLAM algorithm are reduced to avoid complicated structures and execute in a low cost handheld system. In the second step, a rough map is generated by designed handheld system. This rough map called as pre-map. Purpose of the generated pre-map is to use it in other studies, to assist solving localization problem.  Index Terms-pre-mapping, LIDAR, GMAPPING, scan matching, grid base mapping, rao-blackwellized  Manuscript

Implementation of SLAM, Navigation, Obstacle Avoidance, and Path Planning of a Robust Mobile Robot Using 2D Laser Scanner

“…However, glass doors/windows and stairs pose great challenges and may require manual adjustments to modify these areas on the map. This process of mapping is called pre-mapping [22], which serves an important role in better understanding a disaster area. In [22], LIDAR was used to outline this 15 idea, although they faced several problems such as floor plan distortion by a catastrophic event.…”

“…In one particular study [22], a classroom with narrow tables has been chosen to test the Gmapping. The first test involved an ideal classroom with wider objects and walls that were completely covered by cardboard.…”

“…A major evaluation was conducted after all experiments with respect to different SLAM methods in an effort to find the optimal SLAM method. The results were compiled in [22] and their work has proposed pre-mapping as a reliable solution for lack of reliable results in such environments.…”

“…One particular pre-mapping approach that is common among studies is mapping these environments for emergency rescue teams to use in future natural disasters such as an earthquake. This approach is called Disaster Areas Pre-mapping and has found to be helpful for the USAR team to approximate the location of survivors [22]. It is critical to emphasize that currently, the number of places which have been mapped are few which make this approach more theoretical than practical.…”

An Embedded System Design to Build Real-Time 2D Maps for Unknown Indoor Environments

“…The work carried out in recent years on simultaneous localization and mapping (SLAM) has primarily focused on developing algorithms that are real-time implementable that can obtain more accurate and robust maps [23, 24. One of the ways to increase the accuracy of 2D map extractions in unknown environments is to develop sensor fusion applications [25] as considered in the system proposed in this study. It is also important that the position information be updated quickly, so that a robot in the field can determine its own position more accurately.…”