A Method for Detection of Flat Walls in Through-the-Wall SAR Imaging

Abstract

Through-the-wall imaging using wideband synthetic aperture radar (SAR) is a powerful tool that enables seeing through visually opaque walls by providing high-resolution images of objects behind the walls. Previous works focus on detection of static and moving point scatterers behind the walls and methods for enhancing the image by removing the effect of transmission through the walls. Besides imaging of hidden objects, detection of walls and large flat surfaces provides complete map of buildings’ interiors and enables better path planning and treat assessment in rescue and military operations. In the standard high-resolution SAR processing, walls and objects with large surface are imaged as discrete points instead of solid lines and as a result, without prior knowledge about the imaging area, walls may be interpreted as few discrete closely spaced targets. In this letter, a method is presented to discriminate walls and objects with large flat surfaces from other objects. In this approach, instead of focusing the synthetic radar beam on a point on the wall surface, the beam is focused at the location of the image of transmitter with respect to the wall surface considering only specular reflections. This is done by assuming there exists a wall at a distance from the transmitter with a known orientation. This results in an image in polar format in which locations of peaks determine the distance and orientation of the actual wall surfaces inside the imaging area. The method is applied to the measured SAR data and the results exhibit the capability of the method in detection of walls in real scenarios.