Security surveillance radar systems using feature base neural network learning and security surveillance method thereof

摘要

The present invention is a security monitoring radar using a feature point-based neural network learning that can reduce the complexity and memory requirement of a machine learning algorithm by extracting a feature point from a unit Doppler profile of a Doppler radar through a feature point extractor of low complexity and learning it through a neural network. To a system and its security monitoring method, by extracting a plurality of feature points that can be extracted with a low computational load from a unit Doppler profile of a radar spectrogram, thereby reducing memory requirements, lowering computational complexity, and significantly reducing hardware implementation area to achieve light weight and low cost can have an effect.