Kalman Filter Employment in Image Processing

The Kalman filter is a classical algorithm of estimation and control theory. Its use in image processing is not very well known as it is not its typical application area. The paper deals with the presentation and demonstration of selected possibilities of using the Kalman filter in image processing. Particular attention is paid to problems of image noise filtering and blurred image restoration. The contribution presents the reduced update Kalman filter algorithm, that can be used to solve both the tasks. The construction of the image model, which is the necessary first step prior to the application of the algorithm itself, is briefly mentioned too. The described procedures are then implemented in the MATLAB software and the results are presented and discussed in the paper.