criteria for assessing the adequacy of image recognition methods and their verification using examples of artificial series of signals

The article discusses four criteria for assessing the adequacy of the most well-known image recognition methods. verification of two of these criteria is carried out by empirical analysis using the example of the most well-known signal recognition methods, such as dtw, ddtw, as well as methods based on the wavelet transform and fourier transform. two artificial sets of images are used as recognition objects, formed by uniformly shifting the base image both horizontally and vertically. in general, the goal of this research is to develop a new method for extracting recognition features using the example of the image of the state emblem of the republic of azerbaijan. in the context of this study, a verification of a previously proposed signal recognition algorithm is carried out based on the artificial family of curves, for which the most accessible and acceptable method of displacement is established: horizontally or simultaneously horizontally and vertically.