Towards ocean wave spectrum estimation from marine radar data by the polar Fourier transform

In marine radar remote sensing, scan conversion can be visualized as a process of re-sampling in which the original image is recovered and then sampled on a new grid. Recovering an image from its samples implies applying a low pass filter which would present some distortion in the wave spectrum. In this paper, a new approach to acquire the wave spectrum with minimal distortion is proposed. The basic idea is to avoid the whole scan conversion stage and apply a Fourier transform directly on B-scan images. Of course, the regular Fourier transform cannot help here. Therefore, a new Fourier-like transform, referred to as the polar Fourier transform, is adopted in order to generate the power spectrum. The new Fourier transform represents an image in terms of cylindrical waves rather than plane waves as in the regular Fourier transform. Results show the potential of the proposed polar Fourier transform to estimate the wave spectrum without using scan conversion.