Augmented Monte Carlo Localization
alifahrri/robosoccer-mcl as my baseline codes.
Credit toRequirements
1. OpenCV <=4
2. Qt
3. YAML
4. C++11
Build and use using CMake
cmake .
make
./mcl_localization
This is an Augmented Monte Carlo Localization for a humanoid robot with line points observation that is similar to a mobile soccer robot in MSL League.
Approach A. Field Points
The idea is because we don't see the field lines all the time, why don't we observe the white points from a line or boundary of the outer field and then use those points to calculate our position?
Similar to the popular method used in Middle Size League Soccer Robot:
Approach B. Line Segments
Field lines can be detected using many methods and can be used for localization. Here, I explain a simple way to use observed line segments and compare it with known field lines. So we do not need to see the entire line and compare it the refrence line. We only need parts of it as long it's inside the reference line. I forgot the original paper reference.
Self Localize
Field Points | Field Lines |
---|---|
Kidnapping Problem
Simple line scanning to find field points --DISABLED--
See the robot's fov area in white. This is a very simple using line iterator
Field Points | Field Lines |
---|---|
Youtube Video
Control
Key | Action |
---|---|
W | +5 in Robot's X |
S | -5 in Robot's X |
A | -5 in Robot's Y |
D | +5 in Robot's Y |
E | -2 in Robot's Heading |
Q | +2 in Robot's Heading |
Use Heading | Sse robot's heading for weighting |
Adaptive Particle | Use the adaptive particle algorithm |
Debug Features | To view lines from robot fov DISABLED |
Features:
- MCL
- AMCL
- Line Points Observation
- Field Lines Observation
- Number of particles adaptation
Feel free to ask me, I will be very happy to discuss and learn from others. m.haritsah@mail.ugm.ac.id
References
- Huimin Lu, Xun Li, Hui Zhang, Mei Hu & Zhiqiang Zheng (2013) Robust and real-time self-localization based on omnidirectional vision for soccer robots, Advanced Robotics, 27:10, 799-811, DOI: 10.1080/01691864.2013.785473
- Lu, H., Yang, S., Zhang, H.B., & Zheng, Z. (2011). A robust omnidirectional vision sensor for soccer robots. Mechatronics, 21, 373-389.
- Lauer M., Lange S., Riedmiller M. (2006) Calculating the Perfect Match: An Efficient and Accurate Approach for Robot Self-localization. In: Bredenfeld A., Jacoff A., Noda I., Takahashi Y. (eds) RoboCup 2005: Robot Soccer World Cup IX. RoboCup 2005. Lecture Notes in Computer Science, vol 4020. Springer, Berlin, Heidelberg
- Sebastian Thrun, Wolfram Burgard, and Dieter Fox. 2005. Probabilistic Robotics (Intelligent Robotics and Autonomous Agents). The MIT Press.
- Aguiar, L., Máximo, M. and Pinto, S. (n.d.). Monte Carlo Localization for Robocup 3D Soccer Simulation League.
- Messias, J., Santos, J., Estilita, J. and Lima, P. (2008). Monte Carlo Localization Based on Gyrodometry and Line-Detection. In: 8th Conference on Autonomous Robot Systems and Competitions.
- Muzio, A., Aguiar, L., Máximo, M., & Pinto, S.C. (2016). Monte Carlo Localization with Field Lines Observations for Simulated Humanoid Robotic Soccer. 2016 XIII Latin American Robotics Symposium and IV Brazilian Robotics Symposium (LARS/SBR), 334-339.