Toward practical implementation of sensor fusion for a demining robot

This paper presents a pre-processing method to allow the implementation of landmine sensor fusion techniques in a pneumatic demining robot. The proposed method is based on a two-level strategy for the sensor fusion. This strategy allows separating two different classification tasks: object/background and mine/another object. This work is focused in the first stage of the process, particularly in the identification of regions-of-interest (ROI). The proposed ROI extraction algorithm decreases the number of parameters, is reliable in different environmental conditions and can be implemented online. Results of experimental implementation and conclusions are presented.