Digital image processing algorithms for visual navigation of an unmanned aircraft

Abstract

The algorithms of digital image processing in the unmanned aerial vehicle (UAV) visual navigation system are stated and the program realization of the method is presented in the article. Assessment of a range of spatial frequencies of the image is carried out using the stated algorithms. The look and structure of a two-dimensional amplitude spatial spectrum of the image allow allocating existence of vertical and horizontal borders and lines on the image and giving a general idea about their spatial orientation. The spectral submission of the compared images is also used for the acceleration of correlation function calculations. Using the results of digital processing of an amplitude two-dimensional spatial range of the image observed by the onboard TV camera the data allowing to specify the spatial UAV coordinates and to estimate a condition and position of the airfield are formed. The method is based on the numerical analysis of a two-dimensional amplitude spectrum of the image. Results of operation that confirm efficiency of the stated algorithms are presented.