Localization and map building for a mobile robot

In many applications of mobile robotics, it is significant to be able to locate the robot with respect to its environment. We present an approach based on the use of telemetric and odometry data for the mapping and the localization of a mobile robot moving in a structured but unknown environment. The algorithm of fusion between odometry data and US data with the aim of localization implement an Extended Kalman Filter using the odometry data in the equation of evolution and the measurement provided by the US sensors in the equation of observation. A second filter, permits to update, after each localization, an incremental representation of the scene defined in the initial reference frame, in a integrated strategy of perception.