On application of two-dimensional AR modeling of Wigner distribution for efficient coding and effective filtering

The use of two-dimensional AR (autoregressive) modeling of the Wigner distribution (WD) to perform efficient coding of WD images and effective filtering of troublesome cross WD components and white noise components is demonstrated. The AR models are obtained using two-dimensional autocorrelation fitting techniques that are based on nonsymmetric half plane or quarter plane model support. The two-dimensional AR model provides an alternative time-frequency representation which can have improved time and frequency resolution, fewer cross terms, and less white noise interference. Examples demonstrating that two-dimensional AR modeling of the WD reduces the storage computation by one to two orders of magnitude are given