Data interpolation and resampling for a synthetic aperture radar data

In this paper, we introduce interpolation for a detailed observation of synthetic aperture radar (SAR) data. The interpolation is done by zero-padding in the frequency domain. By comparing the raw SAR data and the interpolated data, for example, it is observed that the raw span at a corner reflector we deployed is 0.75 times smaller than its real values. Also, we propose a resampling method applied to the interpolated data to keep data amount same as raw data. The proposed method resamples points where peaks of the span due to coherent targets exist. As for pixels where the peak does not exist, points with the smallest polarimetric entropy are selected. The proposed method is compared to raw data and the resampled data by the maximum filter. It is confirmed that the proposed data emphasizes the characteristics of the raw data. Also, its blurring effect is not as much as that in the resampled data by the maximum filter. This technique is quite basic, however, the effect is not negligible.