Search

In the process of forming and transmitting images through communication channels, they are exposed to various types of noise, which in some cases leads to deterioration of visual quality and loss of areas of images. Reducing such defects is an urgent task in many applications in filtering and recovering data. Methods and technologies of data recovery in digital image filtering problems have been studied, which made it possible to formulate the main directions of improving data recovery processes without deteriorating their quality. Known classical filtering methods, such as the median filter, Gaussian filter, Wiener filter and others, used in known modern image processing software packages, often do not have satisfactory efficiency in processing specific types of noise and often lead to partial loss of useful signal. The aim of the work is to increase the efficiency of image filtering in the development of a new method of the smallest finite differences and the creation of an interactive web service for its implementation. The proposed method of filtering allows you to effectively eliminate pulsed noise from video data, taking into account a priori information about the noise without significant distortion of useful information. The found quantitative characteristics of the evaluation of the efficiency of the method show better results of pulse noise filtering in the image in comparison with the known ones. During the development of the method, a physic-mathematical interpretation of the recovery of digital images when interacting with additive pulse noise was used, and a gradient analysis model was used using the final differences of the pixel values of the images. The proposed algorithm is used to restore test images, the analysis of digital processing in comparison with other known methods using quantitative and qualitative characteristics. The advantage of the created Internet service is the implementation of flexible and efficient functions of digital image filtering, which is absent in the well-known graphics packages Adobe Photoshop, Paint, GIMP and others. In addition, the proposed solutions in the form of an Internet service are a good compromise between ease of implementation and quality of filtering, compared to powerful but difficult to implement and resource-intensive image processing systems. [ABSTRACT FROM AUTHOR]

The results of synphytoindication estimation of parameters of climatic and edaphic factors are presented. They characterize meadow-steppe areas of north-western part of Podillya within the limits of the Lviv and Ivano-Frankivsk region. Differentiative influence of each ecological factor is determined in figuring out the investigated objects’ position in the coordinates’ system of ecotopic features in steppe ecosystems of East-Europe forest-steppe and Prychornomorska (Pontychna) steppe provinces. A certain role of ecological factors in forming the specific meadow steppes’ position of Podillya in the ecological continuum of European part of the Eurasia Steppe region is defined.