Image superimposition technique in computer vision systems using contour analysis methods

Abstract

The works presented by the authors at MECO-2013 - MECO-2015 conferences as well as the other works considered the components of image superimposition techniques in computer vision systems with different degree of detail. In this paper for the first time a complete description of the automatic image superimposition technique is provided. This technique is suitable for the images both of a similar and of a various nature. The results of the implementation of the proposed technique for an image sequence from the aircraft on-board camera are represented. Also for the first time the practical utilization of the complex contour analysis theory to solve the problem of establishing one-to-one correspondence between object sets in the pair of real images for aviation computer vision systems is accomplished.