Adaptation of the Stuff Algorithm to Realistic Measurement Scenarios

Identifying non-cooperative and potentially hostile vehicles is of the utmost importance to the operators of manned and un-manned surveillance aircraft. Utilizing ISAR imagery has been one method for performing target identification by many organizations. However, the complicated motions of the target vehicle and ISAR platform often prevent a clear 2D image from being recovered. Over the course of the previous 10 years, methods have been developed to utilize the complicated target motions to convert 1D range data into a 3D image of the target. The methods, however, have thus far not been able to stand up to application of real data. The presence of scatterers not attached to the rigid target body and scatterers which are not visible for wide view angles both cause difficulties for these methods. The DAC-MTRI team proposes new algorithms and systems which will begin to handle these issues. The approach will create a step forward in creating a continuous link between ISAR measurements, advanced imaging techniques, and target recognition systems. BENEFIT: As the need for automatic target recognition systems increases, the ability to produce reasonable images of target objects increases as well. This work will expand on the methods developed over the previous 10 years which produce 3D images based on 1D ISAR range data. The ability to handle noisy data with non-persistent scatterers is a significant obstacle in the development of these methods. With the systems presented in this work, we plan to overcome many of these difficulties.