A reduced order extended Kalman filter for sequential images containing a moving object
- 1 July 1993
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 2 (3) , 285-295
- https://doi.org/10.1109/83.236537
Abstract
The extended Kalman filter (EKF) is applied to the reduction of noise in sequential images containing a moving object and to the estimation of the object's velocity. A computationally tractable approximation of the EKF, called the parallel extended Kalman filter (PEKF), is generated. The PEKF consists of a parallel bank of third-order EKFs, operating on the Fourier coefficients of the image, followed by a finite impulse response filter. The PEKF is shown to converge to the optimal (in the mean square sense) algorithm in the limit as the velocity estimation errors approach zero. The performance of the PEKF is demonstrated for very low signal-to-noise ratio (SNR) images. The PEKF also provides a natural setting for tracking slow changes in the object (real or apparent) and its velocity, since these variations are included in the model. The relation of the PEKF to another frequency domain algorithm for velocity estimation is discussed. The algorithm is illustrated by application to an example and its performance is demonstrated in the presence of velocity estimation errors.Keywords
This publication has 17 references indexed in Scilit:
- The image flow constraint equationComputer Vision, Graphics, and Image Processing, 1986
- A fast Kalman filter for images degraded by both blur and noiseIEEE Transactions on Acoustics, Speech, and Signal Processing, 1983
- Estimating three-dimensional motion parameters of a rigid planar patchIEEE Transactions on Acoustics, Speech, and Signal Processing, 1981
- Fourier analysis using spectral observersProceedings of the IEEE, 1980
- Determining the movement of objects from a sequence of imagesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1980
- A model for processing of movement in the visual systemBiological Cybernetics, 1979
- A new phase unwrapping algorithmIEEE Transactions on Acoustics, Speech, and Signal Processing, 1977
- Computer Analysis of Moving Polygonal ImagesIEEE Transactions on Computers, 1975
- On the estimation of state variables and parameters for noisy dynamic systemsIEEE Transactions on Automatic Control, 1964
- LINEAR REGRESSION APPLIED TO SYSTEM IDENTIFICATION FOR ADAPTIVE CONTROL SYSTEMSAIAA Journal, 1963