Study on Monte Carlo Localization and Real-time Map Building

碩士 === 國立高雄第一科技大學 === 電機工程研究所碩士班 === 103 === If wanting to work in an unknown environment, then people need to explore and record the environment information. If wanting to work in a known environment, then they need to localize the position of the this environment to execute assigned missions, simi...

Full description

Bibliographic Details
Main Authors: Guang-Cheng Lin, 林廣誠
Other Authors: Kuo-Yang Tu
Format: Others
Language:zh-TW
Published: 2015
Online Access:http://ndltd.ncl.edu.tw/handle/81669617880998413874
Description
Summary:碩士 === 國立高雄第一科技大學 === 電機工程研究所碩士班 === 103 === If wanting to work in an unknown environment, then people need to explore and record the environment information. If wanting to work in a known environment, then they need to localize the position of the this environment to execute assigned missions, similar to humanoid robots. In this experiment, Similar to real human, humanoid robot sensor only uses a computer vision system with webcam is designed for localization and map building, rather than using ultrasound, infrared and laser sensors for distance measurement. To finish the localization and map building depends on single camera is quite challenging for humanoid robot system. In this thesis, Monte Carlo localization method for localization in the known map, and update the map information immediately with unknown obstacles. In addition, for reducing convergence time of Monte Carlo localization, combined with trilateration method to the initial position is also included.