Kalman filter for denoising of infrared images
Ramesh K. Agarwal, Ramesh K. Agarwal · 32nd Thermophysics Conference · 1997
In this paper the use of Kalman filter for iterative denoising of the infrared (IR) images is presented. The denoising model employed is a special linearized version of the nonlinear heat equation. Kalman filter is employed as a noise estimator in the denoising algorithm which provides correction to the contaminated image. The technique is applied to several analytic images. The robustness and accuracy of the method is compared with the nonlinear noise removal algorithm of Rudin, Osher and Fatemi [1] which has been recently implemented by Bihari [2] using the constrained optimization approach. The algorithm of Rudin, Osher and Fatemi minimizes the total variation (TV) of the actual image.