UDC 004.932
COMPARATIVE ANALYSIS OF DISCRETE GAUSSIAN NOISE
FILTRATION METHODS IN IMAGE PROCESSING
A. I. Novikov, Ph.D. associate professor, RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
A. V. Pronkin, student of Faculty of Computer Science RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
This research of the work is dedicated to smooth properties and simultaneously to a degree of border blurring of linear and nonlinear filters while processing images. The aim of this work is generation of practical recommendations for selection of filters that suppress discrete Gaussian noise, suited for work in airborne computer under the conditions of severe limitations to processing speed of algorithms. In airborne systems of computer sight the processing of images must be carried out in real-time mode. For the majority of systems this corresponds to 25 frames per second. The problem of noise suppressing is a low-level one and therefore must be processed much faster than high-level ones. In this article the results of comparative tests of groups of linear and nonlinear filters of discrete Gaussian noise are stated and modification for bilateral filter, permitting to speed up image processing up to 10 times is proposed.
Key words: images, additive model of image, smoothing filter, kernel smoothing operator, linear filters, sigma-filter, bilateral filter, image blur.