Microwave Imaging Using Spatially Separated Radiated Fields

The patent segments the imaging domain into a grid of discrete pixels and separates the scattered field into contributions from individual pixels. By using directional Green's function kernels that isolate radiation from each pixel in specific directions, the method decomposes the complex inverse scattering problem into simpler, independently solvable components, thereby improving measurement precision without proportionally increasing algorithm complexity

The patent introduces directional Green's function kernels as intermediary mathematical tools that mediate between the scattered field measurements and the pixel-wise radiation sources. These kernels act as filters that selectively extract directional information from the scattered field, transforming the ill-posed inverse problem into a more tractable form that achieves higher resolution with reduced computational complexity

The patent replaces physical separation apparatus with mathematical directional Green's function kernels that perform the separation function in the data processing domain. Instead of using complex optical or mechanical separation devices to direct radiated fields to different observation locations, the method uses computationally implemented directional kernels to achieve the same effect, thereby improving problem conditioning without proportionally increasing device complexity

The patent changes the parameter space by transforming the scattered field data through directional Green's function kernels with specific angular parameters. By selecting kernels with appropriate directional characteristics, the method transforms the measurement data into a parameterized form that directly yields a well-conditioned system matrix, enabling reliable image reconstruction without additional physical separation components