Higher-than second-order statistics are used in the paper to derive and implement two classes of inverse filtering criteria for consistent parameter estimation of asymmetric noncausal ARMA image models with known orders. Contrary to existing approaches, FIR inverse filters are employed and as a result image models with zeros on the unit bi-circle can be handled. The performance of these parameter estimators is evaluated with Monte Carlo simulations and texture classification experiments.
|Original language||English (US)|
|Number of pages||5|
|Journal||Proceedings - International Conference on Image Processing, ICIP|
|State||Published - Jan 1 1994|
|Event||Proceedings of the 1994 1st IEEE International Conference on Image Processing. Part 3 (of 3) - Austin, TX, USA|
Duration: Nov 13 1994 → Nov 16 1994