A Rational Approach to the Problem of Accurate UAV Landing Using Intelligent Image Processing Methods

Abstract

This paper is devoted to the description of preliminary results of solving two problems related to the problem of accurate unmanned aerial vehicle (UAV) landing. The first task is to develop a methodology for rational choice of the landing platform image, which ensures the best recognition by an UAV from any angle and a reasonable observation distance. The second task is to develop a procedure for a sequential analysis of the current situation, forthcoming of the UAV to the platform and accurate landing. To solve these problems, it is proposed to use the previously developed intelligent processing methods based on the using of the structural properties of an image. In particular, the technique is applied using the Weibull distribution model for the gradient magnitude of an image and its components. Numerical results are presented that show the prospects of the proposed procedures and the directions for improving the developed techniques.