MODEL-BASED MULTIRATE KALMAN FILTERING APPROACH FOR OPTIMAL 2-DIMENSIONAL SIGNAL RECONSTRUCTION FROM NOISY SUBBAND SYSTEMS

Conventional synthesis filters in subband systems lose their optimalit y when additive noise (due, for example, to signal quantization) distu rbs the subband components. The multichannel representation of subband signals is combined with the statistical model of input signal to der ive the multirate slate-space model for the filter bank system with ad ditive subband noises. Thus the signal reconstruction problem in subba nd systems can be formulated as the process of optimal state estimatio n in the equivalent multirate state-space model. Incorporated with the vector dynamical model, a 2-D multirate state-space model suitable fo r 2-D Kalman filtering is developed. The performance of the proposed 2 -D multirate Kalman filter can be further improved through adaptive se gmentation of the object plane. The object plane is partitioned into d isjoint regions based on their spatial activity, and different vector dynamical models are used to characterize the nonstationary object-pla ne distributions, Finally, computer simulations with the proposed 2-D multirate Kalman filter give favorable results, (C) 1998 Society of Ph ota-Optical Instrumentation Engineers.