Applications of 3-D LCR networks in the design of 3-D recursive filters for processing image sequences

摘要

Applications of three-dimensional (3D) LCR prototype networks in the design of 3D recursive filters are reviewed and several types of new and useful 3D LCR prototype filters are proposed and classified. These 3D LCR prototype filters can be used to design 3D recursive filters for the processing of image sequences involving 3D spatio-temporal-domain linear-trajectory (LT) and planar-wave (PW) component signals. Among the new proposed filters, a 3D planar-rejection (PR) filter can be used to reject an undesired spatially-varying or spatially-static background LT component signal; a 3D planar-pass/planar-rejection (PP-PR) filter can be tuned to enhance a desired 3D LT component signal and, at the same time, reject an undesired 3D LT component signal. A technique for implementing the 3D filter transfer functions is proposed that employs 3D differential or 3D integral operators, leading to highly-parallel and modular 3D recursive filter structures that are efficient in computation. A simple design procedure is described and examples of image sequence processing are given.