NEURAL NETWORK ADAPTIVE SYSTEM FOR RECOGNISING OBJECTS FROM ACOUSTIC EMISSIONS THEREOF Russian patent published in 2014 - IPC G06K19/00 G06F19/00 

FIELD: physics.

SUBSTANCE: neural network adaptive system for recognising objects from acoustic emissions thereof consists of a receiving unit which includes a receiver and a preprocessing unit, having a preamplifier, a unit for selecting information-bearing features, an adaptation unit, a target type recognition unit which is in form of a neural network processing circuit, the system further including an energy channel unit, a switch and a display unit.

EFFECT: wider range of classified military equipment from acoustic emissions thereof using adaptation of a signal processing algorithm to speeds of aerodynamic and ground objects.

2 dwg