China
Africa
Explore chapters and articles related to this topic
Sensors for Autonomous Vehicles in Infrastructure Inspection Applications
Published in Diego Galar, Uday Kumar, Dammika Seneviratne, Robots, Drones, UAVs and UGVs for Operation and Maintenance, 2020
Diego Galar, Uday Kumar, Dammika Seneviratne
Occupancy grid mapping is used for navigation and localization of autonomous vehicles in dynamic environments (see Figure 4.28). Sensor data from cameras and LiDAR are fused. The camera provides high-level 2D information, such as color, intensity, density, and edge information, and LiDAR provides 3D point cloud data. The usual approach to occupancy grid mapping is to independently filter all grid cells. However, the new trend is to use super pixels for the grid map; the grid cells occupied by an obstacle are included (Oh & Kang, 2016).
Navigation, Environment Description, and Map Building
Published in Marina Indri, Roberto Oboe, Mechatronics and Robotics, 2020
Henry Carrillo, Yasir Latif, José A. Castellanos
Underlying all SLAM algorithms is the estimation problem of recovering the robot position and the environmental representation of the world, given the movements commanded to the robot and the measurements taken from the environment. We could also aim to recover the history of robot positions, which has been shown to improve the estimation’s result due to a better conditioning of the problem [21]. Moreover, it allows useful environmental representations for robotic navigation to be maintained as occupancy grid maps.
Map completion from partial observation using the global structure of multiple environmental maps
Published in Advanced Robotics, 2022
Yuki Katsumata, Akinori Kanechika, Akira Taniguchi, Lotfi El Hafi, Yoshinobu Hagiwara, Tadahiro Taniguchi
The occupancy grid map is one of the map representations that robots commonly use to accomplish various service tasks [1,15–17]. It is a method that divides the environment into grid cells at fixed intervals and stores the occupancy probability of each cell in a 2D list format. The cells with a high value of occupancy probability are assumed to be occupied, and the cells with a low value are assumed to be unoccupied and available for the robot to move to. The occupancy grid map is more suitable for robot navigation tasks than other map representation methods because it can search for routes not occupied by obstacles on the map. Assuming that the grid cell of the occupancy grid map with index i is , the occupancy grid map m is the space divided by grid cells such that where S is the number of cells in the occupancy grid map. A binary occupancy value is assigned to each grid : when occupied and when unoccupied. An occupancy grid map is used to expand an uncertain or unspecified region as an unsearched area. The unsearched area is a cell for which it is impossible to determine its occupancy from the sensor values.
Online glass confidence map building using laser rangefinder for mobile robots
Published in Advanced Robotics, 2020
Jun Jiang, Renato Miyagusuku, Atsushi Yamashita, Hajime Asama
Mapping problem: For mapping, a common approach is the use of occupancy grid maps [4]. Occupancy grid maps divide the area to be mapped into cells using a regularly spaced grid. In a nutshell, for each cell, the probability of it being occupied or unoccupied is calculated after every scan. If the sensor sees an object at the cell, it is considered occupied, if an object behind the cell is observed, it is considered unoccupied.
Sparse-Map: automatic topological map creation via unsupervised learning techniques
Published in Advanced Robotics, 2022
Jesús Hernández, Jesús Savage, Marco Negrete, Luis Contreras, Carlos Sarmiento, Oscar Fuentes, Hiroyuki Okada
In general, the most popular type of two-dimensional map representation is the occupancy grid; originally, these were built using SONAR sensors, however, most recent algorithms use LIDAR sensors, leveraging the high sampling rate and accuracy of laser technology. In the case of three-dimensional maps, a generalization of occupancy grid based on octrees named Octomaps [9], allows for a dense 3D representation with reasonable memory usage.