Using the set of informative features of a binding object to construct a decision function by the system of technical vision when localizing mobile robots

Abstract

The object of this study is the process of constructing a decision function by the optical-electronic system of technical vision under the conditions of the influence of obstacles on the current image, which is formed in the process of localization of a mobile robot. The paper reports the results of solving the problem of constructing a decision function when reducing the signal-to-noise ratio of the current image by using a set of informative features for the selection of the binding object, namely, brightness, contrast, and its area. The selection of the binding object is proposed to be carried out by choosing the appropriate values of the quantization thresholds of the current image for the selected informative features, taking into account the signal-to-noise ratio, which provides the necessary probability of object selection. The dependence of the object selection probability on the selected values of the quantization thresholds was established. The use of the results could enable the construction of a unimodal decision function when localizing mobile robots on imaging surfaces with weakly pronounced brightness and contrast characteristics of objects, as well as with their small geometric dimensions. By modeling, the probability of forming a decision function was estimated depending on the degree of noise of the current images. It is shown that the application of the proposed approach allows selection of objects with a probability ranging from 0.78 to 0.99 for the values of the signal-to-noise ratio of images formed by the technical vision system under real conditions. The method to construct a decision function under the influence of interference could be implemented in information processing algorithms used in optical-electronic technical vision systems for the navigation of unmanned aerial vehicles