Вход на сайт

Building of a distorted images fuzzy estimates.

 
Kulishova N. Ie., Popova A. A. (Kharkivskyi natsionalnyi universytet radioelektroniky, Kharkiv, Ukraina) Building of a distorted images fuzzy estimates. — 2014. — № 3(45). — pp. 36–45.
During the preparation of the issue to publication is sometimes necessary to process many graphical elements in the same style, which involves the repetition of the same operations, performing sequences of elementary actions, such as creating a new image, adding, deleting layers, selection areas, the creation of the text, the use of various filters etc. Another problem connected with package of images processing when it is necessary to apply the same transformation directly to a large group of images in the preparation of information and advertising catalogs, websites, portals, shops. The processing complexity rises due to the fact on what criteria the resulting quality of the original images is evaluated.
The quality of such a complex object as image is important, but it is also quite fuzzy concept. Although digital image processing based on mathematical and probabilistic formulations, human intuition and analysis play a central role in the choice of a particular method among others, and this choice is often performed on the basis of subjective visual assessment.
Subjective perception does not allow to apply rigorous mathematical optimality criteria. This paper proposes a formal approach of estimating the degree of noise and artifacts that can reduce the processing time a large number of originals with similar characteristics. This approach is based on the use of linguistic variables, membership functions were built on the basis of expert evaluation procedures.
Keywords: digital images, noise, blur, linguistic variable, the membership function.
Language — Ukraine.
Illustrations — 4.
Bibliography — 13 titles.
PDF