Research On Navigation Of Indoors Autonomous Mobile Robot

The research of autonomous mobile robot's navigation mainly focuses at how to realize a robot's perception,positioning,cognition and movement manipulation in a real environment.This thesis lays special emphasis on research of realizing a robot's patrolling the inside of a building.First,a building's external picture and plane map of its indoors structure are scanned into the computer.Then,we detect three vanishing points from the external picture using sine curve fitting in the parameter space after Canny edge detection and Hough Transform extracting straight lines.So,we can build a 2D referenced frame for the plane map because of invariable vanishing points. Therefore,the thesis deeply investigate how to extract lines using Hough Transform, including the analysis and implementation of a reasonable quantization of the parameter space and the determination of line segment's endpoints.Last but not least, we measure and decompose the plane map in order to represent the indoors environment of the building and compute a path planning so that simulate a robot's patrolling in the plane map.