Improving a method for segmenting optical-electronic images acquired from space observation systems based on the firefly algorithm

The object of research is the process segmentation optoelectronic images acquired from space observation systems. method systems based on firefly algorithm, unlike known ones, involves following: – pre-selection brightness channels Red-Green-Blue color in original image; calculation level luminosity for each firefly; assigning with neighboring firefly, within a certain radius, whose higher than natural determination coordinate updated position channel. An experimental study into algorithm was carried out. It established that improved allows quality by evaluated comparison methods particle swarm and Sine-Cosine algorithm. found reduces error first kind an average 11 % second 9 %. This becomes possible using Methods image can be implemented software hardware processing surveillance Further studies may focus comparing genetic algorithms.