An unmanned aerial vehicles navigation system on the basis of pattern recognition applications—Review of implementation options and prospects for development

Abstract

The popularity of unmanned aerial vehicles (UAVs) kept growing over the past few decades. UAVs are widely used in various fields for monitoring, mapping, aerial photography, rescue operations, and so forth. However, the UAV navigation may be difficult in areas where global positioning systems is not available. This article presents an analytical review of the promising methods of creating a navigation system based on a pattern recognition algorithm. Linear algebra and mathematical morphology to describe the operation of pattern recognition applications in UAV s navigation systems, classification of images through profound research of intelligent navigation systems, analysis and digital image processing to identify options for the implementation of artificial intelligent systems of visual navigation were used to identify existing pattern recognition applications in the process of constructive systematization. It was shown that the UAV navigation requires objects in the image to be recognized and the distance to a possible obstacle to be estimated using deep learning method for image segmentation and depth map estimation. An obligatory initial point in the image recognition algorithm is the image preprocessing procedures based on machine vision techniques. The results of this study may be useful in developing efficient UAV applications for military and civil purposes.