Probabilistic assessment of objects belonging to a given class based on the combination of multispectral images of different times

Object detection is the task of classifying and localization of objects in an image. This article discusses methods of image aggregation. Probabilistic indicators of the quality of optoelectronic means when detecting images of observed objects are determined. The technique of image aggregation in a multispectral optoelectronic system based on the probabilistic belonging of an object to a given class is presented. Probabilistic data on the object belonging to a given class using the machine learning method were obtained. The results of image aggregation with different parameters of the classification criterion are presented.