LAUSR

Radar holography has been established as an effective image reconstruction process by which the measured diffraction pattern across an aperture provides information about a threedimensional target scene of interest. In general, the sampling and reconstruction of radar holographic images are computationally expensive. Imaging can be made more efficient with the use of sparse sampling techniques and appropriate interpolation algorithms. Through extensive simulation and experimentation, we show that simple interpolation of sparsely-sampled target scenes provides a quick and reliable approach to reconstruct sparse datasets for accurate image reconstruction leading to reliable concealed target detection and recognition. For scanning radar applications, data collection time can be drastically reduced through application of sparse sampling. This reduced scan time will typically benefit a real-time system by allowing improvements in processing speed and timeliness of decision-making algorithms. An added advantage is the reduction of required data storage. Experimental holographic data are sparsely sampled over a two-dimensional aperture and reconstructed using numerical interpolation techniques. Extensive experimental evaluation of this new technique of interpolation-based sparse sampling strategies suggests that reduced sampling rates do not degrade the objective quality of holograms of concealed objects.