A Synthetic Aperture Processing for Detecting Underground Cavities

In this paper, we have proposed a fast and accurate method to detect underground cavities in low signal to noise ratio (SNR) situations. The proposed method firstly generates synthetic aperture radar (SAR) imagery from the raw data by Omega-K, which is probably termed Stolt's F-K migration based on stationary phase method. The proposed method divides SAR imagery on azimuth frequency domains as sub-apertures. Since the sub-apertures have characteristics of different incident angles, detecting peaks on every sub-aperture renders the boundary of cavities. The proposed method estimates the accurate boundary position, amplitude and phase of cavities by summing the detected peaks on every sub-aperture. Finally, simulation on 2-dimensional finite-difference time-domain (FDTD) has demonstrated the effectiveness of the proposed method.