Sensor data fusion based position estimation techniques in mobile robot navigation

This paper gives an overview about the position estimation techniques based on typical measurement devices used in mobile robot applications. The purpose of this paper is to give an overview of the position estimation based on the fusion of information from several sensors. It presents different extensions of Kalman filter estimators and analyses the performances of these algorithms. There are compared several estimation techniques like the Extended or Unscented Kalman filters and the particle methods. Furthermore modelling details and stereo vision algorithms are introduced. In the second part there are shown the results of the odometric, ultrasonic measurements techniques and the ones based on stereo vision.