Space/Spatial-Frequency Filtering Based on the Local Frequency Estimation: From the Noisy Signal, through the Space/Spatial-Frequency Domain to the Estimated one

Space/spatial frequency filtering, as a possibility for estimation of two-dimensional highly nonstationary noisy signals, is considered in this paper. Principles of transferring from two-dimensional (2D) to 4D space/spatial-frequency domain, of the local frequency (LF) estimation, and the 2D signal estimation based on the estimated LFs are considered in detail. The filter, based on the considered principles, is tested and verified on a real-life nonstationary noisy 2D signal.