Devising a method for segmenting complex structured images acquired from space observation systems based on the particle swarm algorithm

This paper considers the improved method for segmenting complex structured images acquired from space observation systems based on particle swarm algorithm. Unlike known ones, algorithm involves following: – highlighting brightness channels in Red-Green-Blue color space; using a image each channel of RGB segmentation is reduced to calculating objective function, moving speed, and new location channel. Experimental studies have been conducted by It was established that makes it possible segment surveillance systems. A comparison quality carried out. The comparative visual analysis well-known methods indicates highlights more objects interest (objects military equipment); k-means assigns some (especially those partially covered with snow) snow cover (marked blue); also associates are almost completely blue). has reduces errors first kind an average 12 % second 8