High-Resolution 3-D Imaging Algorithm With an Envelope of Modified Spheres for UWB Through-the-Wall Radars

Abstract

Through-the-wall imaging techniques with ultrawideband (UWB) radars are promising candidates for non-destructive testing and reliable human detection, especially in disaster areas, where victims are buried under collapsed walls. These applications require high-resolution target imaging to identify the object shape, such as a human body. We have already proposed a high-quality 3-dimensional (3-D) imaging algorithm in the form of envelope that is aimed at near field sensing for non-contact measurement or target identification for robots. Envelope achieves real-time accurate 3-D imaging with group mapping from multiple observed ranges to target points, and offers a reliable image even in noisy situations. However, this method does not maintain its quality for through-the-wall imaging because an observed range shift due to wall penetration causes a serious distortion in the image. This paper presents a high-resolution 3-D imaging algorithm by modifying the original envelope, and which gives a more accurate object shape behind a wall. Furthermore, to enhance the resolution of the estimated images, this method is combined with a direct waveform compensation method, known as spectrum offset correction. Numerical simulations and an experiment verify that our proposed method achieves high-resolution 3-D imaging for through-the-wall radar applications.