Sub-band representation image analysis in the field of spatial frequencies

Currently, two-dimensional visual representations of various information arrays are widespread, since images are the most natural form of information exchange for humans. Therefore, various information technologies for the implementation of computer image processing have been created. A signifcant place among existing information technologies is occupied by computer image analysis, the basis of which is the procedure for determining certain properties that characterize them from certain positions. In particular, an important area of image analysis is the automatic classifcation of their constituent objects (pattern recognition). The main attention is paid to the choice of the so-called feature space, which from the standpoint of the problem being solved most adequately refects the properties of the analyzed images. The article considers the possibility of using a sub-band method for image analysis, the application of which, as shown by the research results, allows to obtain characteristics that can be used as features for their comparison.