Using Entropy Estimation to Detect Moving Objects

The article is devoted to the problems of analysis of the informative part of acoustic signals in the UAV control system. A method of determining the frequency range of the informative part of a signal based on the estimation of the normalized energy spectrum and spectral entropy is proposed. The sequence of steps for determining the range of the informative part of the signal, consisting of the calculation of the spectrum of the signal and the normalized energy spectrum, its digital processing, the calculation of the entropy of the spectrum, the median smoothing, logic-time processing. The results of experimental acoustic signal processing for different frame length values are analyzed. Graphic representations of the acoustic signal, spectral entropy values and logic function are given.