Title :
Recursive enhancement of noncausal images
Author :
Balram, Nikhil ; Moura, José M F
Author_Institution :
Dept. of Electr. & Comput. Eng., Carnegie Mellon Univ., Pittsburgh, PA, USA
Abstract :
A recursive procedure is implemented for the enhancement of noncausal Gauss Markov random fields. Experimental results for the enhancement of synthetic as well as real images corrupted by additive white Gaussian noise are provided. These are contrasted with equivalent results obtained by processing the images with recursive filters derived by imposing a causality constraint upon the fields. The results show that the noncausal recursors provide considerable reduction in the mean square error (MSE) of the noisy images without the introduction of undesirable visual effects, such as streaking, that are produced when causality constraints are imposed
Keywords :
Markov processes; picture processing; white noise; additive white Gaussian noise; causality constraint; image processing; mean square error; noisy images; noncausal Gauss Markov random fields; noncausal images; noncausal recursors; recursive filters; recursive image enhancement; Additive white noise; Covariance matrix; Filters; Gaussian processes; Lattices; Markov random fields; Maximum likelihood estimation; Mean square error methods; Noise reduction; Visual effects;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 International Conference on
Conference_Location :
Toronto, Ont.
Print_ISBN :
0-7803-0003-3
DOI :
10.1109/ICASSP.1991.151033
