The SEABED method, a fast imaging algorithm for UWB pulse radar systems, makes it possible to image 3-D unknown targets using a reversible transform between the target shape and antenna position. The antenna position and distance form a curved surface called a "quasi-wavefront". We can estimate the target shape directly using this transform if we can extract the quasi-wavefront from the observed data. In the case of complex-shaped targets, however, it is difficult to obtain the correct image because the received signals interfere with each other through the echoes from multiple reflected points. Conventional methods, that extract the quasi-wavefront at each antenna position, suffer from a serious fault in that the values obtained in the region with interference are not accurate. To overcome this problem, we propose a simultaneous estimation of all quasi-wavefronts from the received signal waveforms at all antenna positions by treating the problem as an optimization of an evaluation
展开▼