Onsets for the Recognition of Objects and Image Refinement using UWB Radar

In a security application a scenario with UWB sensor nodes is given in which the sensor nodes perform autonomous localization, navigation and object recognition tasks. The object recognition is performed by bistatic sensor nodes on the basis of RCS measurements. The corresponding onsets for the recognition process and the tool used in the simulation-based studies are described together with its properties. To study the basic effects several objects were subjected to performance evaluations with regard to object form and robustness of the recognition process. The results show the feasibility but also the need for improvements with regard to the robustness of the moment invariant algorithm.