Environment Modeling

Ultrasonic Sensor Based Grid Mapping

Researchers
- Yun-Kyu Choi (M.S. bazzi83@korea.ac.kr), Hye-Min Han (M.S. hyemin.han@doosan.com)
Research Period
- 2009. 6. ~ 2012. 2.
Research Backgrounds
- Mapping is one of the most important factors for dependable navigation.
- Grid map building based on Bayesian update formula using ultrasonic sensors has been employed.
- Deals with poor angular information (due to wide beam width) and specular reflection of ultrasonic sensors.

Specular reflection of ultrasonic sensor : (a) detection of object(ρ=0),

(b) detection of object (ρ<ω/2), and (c) specular reflection (ρ>ω/2)

Research Objectives
- Grid map building using cheap range sensors
- Geometric data association with accumulated ultrasonic data
- Assignment of the geometric reliability factor of ultrasonic data into the sensor model to build a high-quality grid map.
Research Outputs

Map building in large environment : (a) experimental environment, (b) laser map, (c) sonar map without

consideration of sonar reliability, and (c) sonar map with consideration of sonar reliability.

Robust map building in narrow environments based on combination of sonar and IR sensors

Researchers
- Hye-Min Han (M.S. hyemin.han@doosan.com)
Period of research
- 2010. 4. ~ 2010. 12.
Research Backgrounds
- Poor angular information (due to wide beam width) of the ultrasonic sensor → difficulty in sensing in narrow environments.

Research Objectives
- Grid map building based on combination of low-cost range sensors having different characteristics
- Improvement of mapping accuracy in narrow environments.
Research Output
- The experimental results show that the proposed method can represent the environment such as narrow corridor and open door more accurately than conventional sonar sensor-based map building methods.

* Last updated: 2012. 4. 24