Motivation for Mapping
- Insufficient GPS accuracy
- No GPS indoors
- Random cleaning patterns of roomba
Background to Mapping
- Aspects to mapping
- Localisation
- Mapping
- Simultaneous Localisation and Mapping SLAM
- Type of maps
- Geometric mapping
- Topological mapping
Sliding scale of autonomy Teleoperation Way point/pose navigation supervised autonomous exploration Full autonomy
Map representations
Occupancy Grids
- Map updates
- log likelihood update
Odometry/Dead reckoning
Normally insufficient on its own for map building
Good for providing high speed initial guess
- Sources of odometry
- wheel encoders
- IMU, gyroscopes
- Combination of IMU and wheel encoders can be good
Scan matching
Iterative Closest Point
ICP derivation
Derivation of point-to-plane minimization by Szymon Rusinkiewicz ICP derivation
References Generalized ICP by A. Segal, D. Haehnel and S. Thrun
Mathematics Test
System Message: ERROR/3 (MobileRobotMapping.rst, line 54)
Unknown directive type "math".
.. math:: x^2 + 4
some inline mathematics ( x^2 + 4 )
Using the math directive
Current research
Bobcat Demo
This is an autonomous bobcat
- Localisation by RTK GPS
- Volumetric point cloud map
RANSAC
Random Sample Consensus
System Message: ERROR/3 (MobileRobotMapping.rst, line 83)
Unknown directive type "code-block".
.. code-block:: python
for i in range(10):
print i
References
Simultaneous Localisation and Mapping (SLAM): Part I The Essential Algorithms Hugh Durrant-Whyte, Fellow, IEEE, and Tim Bailey
CS 287: Advanced Robotics, Fall 2009
http://www.cs.berkeley.edu/~pabbeel/cs287-fa09/#syllabus
Lecture on State estimation:mapping
http://www.cs.berkeley.edu/~pabbeel/cs287-fa11/#syllabus