Tracking and enhancement of objects in image sequences using 3-D frequency planar combined DFT/LDE filters

Abstract

Three-dimensional (3-D) frequency planar (FP) adaptive recursive filters have been proposed for tracking smooth trajectory dynamic objects in noisy image sequences. It is shown that the recently proposed class of combined DFT/LDE filters is especially suitable for realizing 3-D FP filters. They are shown that the recently proposed class of combined DFT/LDE filters is especially suitable for realizing 3-D FP filters. They are shown to be capable of recovering linear trajectory (LT) objects from signals containing similar objects having like trajectories, regardless of the spatial size or shape of the object. An adaptive 3-D FP combined DFT/LDE filter is proposed for capturing and tracking objects having nonlinear-discontinuous trajectories and embedded in noise. >