Through-wall ultra-wideband radar human body respiration signal detection method based on Faster-RCNN network

Abstract

The invention discloses a through-wall ultra-wideband radar human body respiration signal detection method based on a Faster-RCNN network. The method comprises the following steps: performing denoising and weak signal enhancement preprocessing on an original slow time-distance two-dimension ultra-wideband radar echo signal; and then continuously performing signal enhancement and denoising on the processed echo by adopting an Advance Normalisation method; imaging the echo signal in the preprocessing process as a gray image by adopting MATLAB, wherein the image width is corresponding to the echoslow time direction, and the image height is corresponding to the echo distance direction; and screening the obtained gray image, annotating a data set, and training a Faster-RCNN network model to perform human body respiration weak signal identification. The method disclosed by the invention is high in identification accuracy rate, high in detection speed, and capable of providing better technical support for the radar echo target detection, the life detection and practical technologies in actual 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 through-wall ultra-wideband radar based on a Faster-RCNN network. Background technique [0002] Ultra-wideband UWB: Ultra Wideband radar life electromagnetic detection uses ultra-wideband electromagnetic waves to detect micro-motion signs such as breathing or heartbeat. In recent years, it has been more and more widely used in actual post-earthquake rescue or safety detection of disaster environments. However, the human body sign micro-motion signal in the radar echo is generally a weak, narrow-band, quasi-periodic signal, and the signal is easily interfered. The time-varying and spatial-varying characteristics of the disaster relief scene and the diversity of noise causes also make Human target recognition has become very complicated, and radar signal recognition technology based on low-dimensional feature...

AI Technical Summary

Problems solved by technology

These methods can effectively remove non-static clutter interference, and successfully extract the number of living entities and their corresponding position and frequency information with high probability and high automation, but they rely heavily on denoising and enhancement algorithms, and weaker life Unrecognizable when signal is masked

Images