digital image processing
E.A. Semenishchev, V.I. Marchuk, I.A. Toropov, I.V. Tolstova
Automated control systems are widely used in various applications of science and technology, such as space research, the analysis of two-dimensional maps of the spatial distribution of signals characteristic of invisible physical fields to identify informative indicators, biological and living systems, the study of growth, development and the dynamics occurring in living organism; medicine for identifying abnormalities in radiographs, thermograms, images of radioisotope and ultrasound, nuclear power, with control of manipulators in areas critical to humans; automation, creating complex systems control objects out of the reach of the operator stations, Oceanography and Marine, in the preparation and research and bottom sediments, increasing the contrast and clarity of images, electron microscope study and industrial X-ray units, topography and lithography, the study of two-dimensional signals received from space, automatic analysis of the terrain, the study of natural resources multispectral television systems, as well as many other areas.
In these systems, an important role is occupied techniques and digital signal processing devices. Image that can be viewed as a two-dimensional signal, is much more capacious storage medium than the usual one-dimensional signal.
The problem of detection and clarify the boundaries of the image appears as a separate problem, after which the solutions it is possible to identify areas with high detail of objects and background, not carrying live traffic.
The methods selection of highly detailed objects in the image on the basis of the relationship means "white / black", "conversions" and "density". The results obtained on a set of test images show the high efficiency of the proposed algorithms, with a small amount of computational effort. These results will allow further processing to produce localized regions of the methods that require high computational cost, but it is accurate and of good quality, as well as improve the overall performance of computer systems.