MODELING AND NONLINEAR FILTERING OF NONSTATIONARY IMAGES |
| 4 | |
| 2010 |
| scientific article | 519.217:004.932:621.396 | ||
| 73-78 | nonstationary digital halftone images, nonlinear filtering, white Gaussian noise, mathematical models of images, Markov chains, final probabilities, transient process, bit binary images |
| The problem of restoration (filtering) of nonstationary digital halftone images destroyed by a white Gaussian noise
is considered. In the image processing, the digital g-bit grayscale image is represented by a set of g binary images.
As a mathematical model of a binary digital image, it is proposed to use a two-dimensional Markov chain with two
unequal-probability initial values, which is a superposition of two one-dimensional nonstationary Markov chains.
A restoration algorithm for such images has been proposed. Computer simulation results of the filtration process are
presented. |
 |