The article describes the method for providing target detection by electro-optical systems with dynamic spectral filtering with difference of spectral characteristics of the current and the reference background signals. The method involves the use of adaptive algorithms and is based on the fact that at a certain value of mismatch between signals the system begins to register the mode of the background spectral characteristics. The normalized Kullback-Leibler divergence is used as the capacity of matching measure of optimum processing in the presence of the spectral variability of input and reference signals. There has been developed a block diagram of an electro-optical system with dynamic control of the spectral characteristics of optical radiation based on acoustic optical filters.
Kullback-Leibler divergence, dynamic spectral filtering, optimal processing of optical signals