An Algorithm for Evaluating the Spectral Characteristics of Regular Textures

摘要

An algorithm for evaluating the spectral characteristics of regular textures on digital images, which are the parameters of the grid of the local maximums of the power spectrum, is proposed. The search and selection of peaks is carried out using the nonlinear spectrum-filtering procedure of an image in polar coordinates. It is shown that the results of the proposed algorithm can be used for solving problems of texture segmentation, detection of texture defects, and filtering of periodic noise.