Abstract: A multistatic array topology and image reconstruction process for fast 3D near field microwave imaging are presented. Together, the techniques allow for hardware efficient realization of an electrically large aperture and video-rate image reconstruction. The array topology samples the scene on a regular grid of phase centers, using a tiling of multistatic arrays. Following a multistatic-to-monostatic correction, the sampled data can then be processed with the well-known and highly efficient monostatic Fast Fourier Transform (FFT) imaging algorithm. In this work, the approach is described and validated experimentally with the formation of high quality microwave images. The scheme is more than two orders of magnitude more computationally efficient than the backprojection method. In fact, it is so efficient that a cluster of four commercial off-the-shelf (COTS) graphical processing units (GPUs) can render a 3D image of a human-sized scene in 0.048-0.101 seconds.

Abstract: A multistatic array topology and image reconstruction process for fast 3D near field microwave imaging are presented. Together, the techniques allow for hardware efficient realization of an electrically large aperture and video-rate image reconstruction. The array topology samples the scene on a regular grid of phase centers, using a tiling of multistatic arrays. Following a multistatic-to-monostatic correction, the sampled data can then be processed with the well-known and highly efficient monostatic Fast Fourier Transform (FFT) imaging algorithm. In this work, the approach is described and validated experimentally with the formation of high quality microwave images. The scheme is more than two orders of magnitude more computationally efficient than the backprojection method. In fact, it is so efficient that a cluster of four commercial off-the-shelf (COTS) graphical processing units (GPUs) can render a 3D image of a human-sized scene in 0.048-0.101 seconds.

Abstract: A multistatic array topology and image reconstruction process for fast 3D near field microwave imaging are presented. Together, the techniques allow for hardware efficient realization of an electrically large aperture and video-rate image reconstruction. The array topology samples the scene on a regular grid of phase centers, using a tiling of multistatic arrays. Following a multistatic-to-monostatic correction, the sampled data can then be processed with the well-known and highly efficient monostatic Fast Fourier Transform (FFT) imaging algorithm. In this work, the approach is described and validated experimentally with the formation of high quality microwave images. The scheme is more than two orders of magnitude more computationally efficient than the backprojection method. In fact, it is so efficient that a cluster of four commercial off-the-shelf (COTS) graphical processing units (GPUs) can render a 3D image of a human-sized scene in 0.048-0.101 seconds.

Abstract: A multistatic array topology and image reconstruction process for fast 3D near field microwave imaging are presented. Together, the techniques allow for hardware efficient realization of an electrically large aperture and video-rate image reconstruction. The array topology samples the scene on a regular grid of phase centers, using a tiling of multistatic arrays. Following a multistatic-to-monostatic correction, the sampled data can then be processed with the well-known and highly efficient monostatic Fast Fourier Transform (FFT) imaging algorithm. In this work, the approach is described and validated experimentally with the formation of high quality microwave images. The scheme is more than two orders of magnitude more computationally efficient than the backprojection method. In fact, it is so efficient that a cluster of four commercial off-the-shelf (COTS) graphical processing units (GPUs) can render a 3D image of a human-sized scene in 0.048-0.101 seconds.