Use of image warping techniques to correct and interpret infrared sensor data for navigation

This paper proposes a method in which sensor fusion techniques coupled with image warping algorithms are used to interpret the signals from two infrared (IR) sensors to relate measures of distance and orientation relative to an obstacle. Development of the algorithm involves modelling of the sensor responses using a set of polynomials. The models are used to develop a 'contour' model which yields a closed-form expression for warping of the sensor levels into a distance-angle plane. The bounds of operation of the algorithm are dictated by the accuracy of the modelling process.