Autonomous Mobile Robot Indoor Environment Modeling

It is not only a process of world-percepting of the autonomous mobile robot modeling, but aslo a prerequisite work of a series of robot location which maybe used to exploration.This article combined the information of the laser sensor with the vision sensor to establish the model of the indoor environment. First it applied the laser sensor as the measurement to obtain the dispositions of the environment, using the arithmetic Lu&Milios to construct the 2D grid map which was based on PIONEER 2-DX8, by this way it successfully avoid either the pertinence problem of the arithmetic Kalman and the disadvantage of the non-real time problem of the arithmetic EM. Data from experiments suggested that it might be more effective when it is confined to a small-ranged environment.In order to increase the varities of the path planing, it is necessary to mark out the positions of doors on the grid map. The article chose the arithmetic Zhang to give vision marks for the camera Canon VC-C4 PTZ, then used the arithmetic of inverse perspective mapping to get information of depth about the monocular visual figure, and made use of the vision sensor to pick up and distinguish the color features of the door. At the same time , it used the method of edge-detect to get the position of the door which is combined with the data from laser sensor, and finally marked out the positions of the doors on the grid map that formed by the laser.