On particle filter localization and mapping for Nao robot
Uwaoma, Andreas Emenike (2015)
Uwaoma, Andreas Emenike
2015
Master's Degree Programme in Machine Automation
Teknisten tieteiden tiedekunta - Faculty of Engineering Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2015-08-12
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201507291492
https://urn.fi/URN:NBN:fi:tty-201507291492
Tiivistelmä
The performance of autonomous mobile robots within an indoor environment relies on an effective detection and localization system. Self-Localization within an indoor environment has been studied and tested experimentally on humanoid robot Nao. The solution utilizes a pre-existing map with known and unknown features.
The aim of this thesis is to utilize map of visual features and the Monte-Carlo Scheme (particle filters) in localization and navigation. Nao robot cameras has been used for detection Naomarks, the detection of these features provides an estimation of the relative distances of features to current robot position. These measurements are applied to a visual localization algorithm that uses a pair of known feature to localize the robot, furthermore the measurements is fused to a particle filter algorithm for estimating the pose of the robot within the map. The particle filter implementation was based on the C++ programming language. A simple path planning scheme was implemented for continuous localization while navigating a paths with obstacles.
The algorithms has been tested with reference to measurements provided by an external sensor. The results of the implementations indicates that the robot can effectively navigate from a start position to a predefined location while avoiding obstacles on its path.
The aim of this thesis is to utilize map of visual features and the Monte-Carlo Scheme (particle filters) in localization and navigation. Nao robot cameras has been used for detection Naomarks, the detection of these features provides an estimation of the relative distances of features to current robot position. These measurements are applied to a visual localization algorithm that uses a pair of known feature to localize the robot, furthermore the measurements is fused to a particle filter algorithm for estimating the pose of the robot within the map. The particle filter implementation was based on the C++ programming language. A simple path planning scheme was implemented for continuous localization while navigating a paths with obstacles.
The algorithms has been tested with reference to measurements provided by an external sensor. The results of the implementations indicates that the robot can effectively navigate from a start position to a predefined location while avoiding obstacles on its path.