A Method for Registration of 3-D Shapes
IEEE Transactions on Pattern Analysis and Machine Intelligence - Special issue on interpretation of 3-D scenes—part II
Globally Consistent Range Scan Alignment for Environment Mapping
Autonomous Robots
Using EM to Learn 3D Models of Indoor Environments with Mobile Robots
ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
An Experimental System for Incremental Environment Modelling by an Autonomous Mobile Robot
The First International Symposium on Experimental Robotics I
An Automated Method for Large-Scale, Ground-Based City Model Acquisition
International Journal of Computer Vision
AAAI'04 Proceedings of the 19th national conference on Artifical intelligence
An Efficient Extension to Elevation Maps for Outdoor Terrain Mapping and Loop Closing
International Journal of Robotics Research
Globally consistent 3D mapping with scan matching
Robotics and Autonomous Systems
Summarizing Image/Surface Registration for 6DOF Robot/Camera Pose Estimation
IbPRIA '07 Proceedings of the 3rd Iberian conference on Pattern Recognition and Image Analysis, Part II
Interest Point Detectors for Visual SLAM
Current Topics in Artificial Intelligence
Multi-robot visual SLAM using a Rao-Blackwellized particle filter
Robotics and Autonomous Systems
Multi-robot map alignment in visual SLAM
WSEAS TRANSACTIONS on SYSTEMS
Map fusion in an independent multi-robot approach
WSEAS TRANSACTIONS on SYSTEMS
Three-dimensional iterative closest point-based outdoor SLAM using terrain classification
Intelligent Service Robotics
Towards multi-robot independent visual SLAM
ICS'10 Proceedings of the 14th WSEAS international conference on Systems: part of the 14th WSEAS CSCC multiconference - Volume I
RGB-D mapping: Using Kinect-style depth cameras for dense 3D modeling of indoor environments
International Journal of Robotics Research
| Hi-index | 0.00 |
Recently, the problem of learning volumetric maps from three-dimenisional range data has become quite popular in the context of mobile robotics. One of the key challenges in this context is to reduce the overall amount of data. The smaller the namber of data points, however, the fewer information is available to register the scans and to conputer a consistent map. In this paper we present a novel approach that estimates global constaints from the data and utilizes these contraints to improve the registration process. In our current system we simultaneously minimize the distance between scans and the distance of edges from planes extracted from the edges to obtain highly accurate three-dimensional modele of the environment. Several experiments carried out in simulation as well as with three-dimensional data obtained with a mobile robot in an outdoor environment we show that our approach yields seriously more accurate models compared to a standard apporach that does not utilize the global constraints.