Human respiratory signal detection method based on faster-rcnn network for through-wall ultra-wideband radar

Abstract

The invention discloses a method for detecting human breathing signals of through-wall ultra-wideband radar based on Faster-RCNN network. Firstly, the original slow time-distance two-dimensional ultra-wideband radar echo signal is preprocessed by denoising and weak signal enhancement; The Advance Normalization method continues to perform signal enhancement and denoising on the processed echo; use MATLAB to image the echo signal in the preprocessing process into a grayscale image, the image width corresponds to the echo slow time direction, and the image height corresponds to the echo distance direction ;Finally, the obtained grayscale images are screened, the data set is marked, and the Faster-RCNN network model is trained to recognize the weak signal of human breathing. The invention has high recognition accuracy and fast detection speed, and provides better technical support for practical technologies such as radar echo target detection and life detection in practical applications.

Description

technical field [0001] The invention belongs to the technical field of radar signal detection, and in particular relates to a method for detecting human breathing signals of a wall-penetrating ultra-wideband radar based on a Faster-RCNN network. Background technique [0002] Ultra Wideband (UWB: Ultra Wideband) radar life electromagnetic detection detects micro-movement signs such as breathing or heartbeat by emitting ultra-wideband electromagnetic waves. In recent years, it has been more and more widely used in actual post-earthquake rescue or safety detection disaster environments. However, the micro-movement signal of human body signs in the radar echo is generally a weak, narrow-band, quasi-periodic signal, which is easily disturbed. The time-varying and space-varying characteristics of the disaster relief site environment, and the diversity of noise causes Human target recognition has become very complex, and it is difficult for radar signal recognition technology based...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a method for detecting human respiratory signals based on the Faster-RCNN network through the wall ultra-wideband radar, aiming at the problem of radar echo human micro-movement signal feature extraction. Apply basic denoising and enhancement algorithms to preprocess the through-wall ultra-wideband radar echo signal, and then apply a deep neural network to extract and identify the weak signal of human breathing in the echo

Images