When operating helicopter-type unmanned aerial vehicles (multicopters) under active operation of counteraction systems, one of the necessary condition for their functioning is the presence of the onboard computer vision system. The use of technical vision systems under disconnection of a multicopter from the GPS signal is accompanied by the use of methods for recognizing images of the Earth’s surface recorded by an onboard camera along the flight path. The variety of image recognition methods existing today dictates the need to choose the most effective recognition method in terms of accuracy. The paper discusses the most popular methods for recognizing continuous signals into which color images of the Earth’s surface are transformed after performing a number of necessary procedures. The image recognition procedure is influenced by a number of factors, such as brightness, presence of noise, viewing angle, etc., as well as methods for converting a RGB model into shades of gray. These factors are due to the impossibility of providing stable shooting conditions. Therefore, it is necessary to compare these recognition methods taking into account their ability to effectively perform their task in these conditions.