作者: Ay?in Ert�z�n , Ahmet H. Kayran , Erdal Panayirci
DOI: 10.1007/BF01260333
关键词: Mathematics 、 Autoregressive model 、 Mathematical analysis 、 Geometry 、 Reflectivity 、 Lattice phase equaliser 、 Entropy (information theory) 、 Mean squared prediction error
摘要: A new lattice filter structure to model two-dimensional (2-D) autoregressive (AR) fields is proposed. The proposed utilizes and extracts the information contained in backward prediction error their delayed versions. main idea use two sets of reflection coefficients corresponding quadrant filters increase number with order filter. Increasing at each stage produces a sufficient independent parameters AR up three, which an improvement over existing 2-D structures. confirmed by computer simulations. In addition, relationship between derived. It also shown that entropy field vector closer input when compared those filters.