A Human Sensing Method Based on 3D Imaging of Outward Ring Scanning Array csar

Abstract

The invention discloses a human body attitude perception method based on the three-dimensional imaging of the outer ring scanning array CSAR. The invention intends to meet the needs of real-time personnel posture perception in nursing homes, smart homes and other places, and install a plurality of circular tracks with CSAR working mode radar arrays to scan outwards to form a non-contact three-dimensional human posture perception system. The array rotates uniformly at a certain angular velocity to form an azimuth synthetic aperture to obtain the target azimuth scattering information, transmits the LFM signal bandwidth through the antenna to obtain the target range scattering information, and uses the height linear array to form a real aperture to obtain the target height scattering information. The range, azimuth, and height scattering information is mapped to Cartesian Cartesian coordinate system, and the 3D scattering information image of the human body in the detected area can be reconstructed, and then the multi-view 3D fusion of scattering information can be performed by using the imaging results of different viewing angles, combined with deep learning. , Image detection and recognition technology, realize 360° non-contact three-dimensional perception of human body posture in the monitoring area.

Description

technical field [0001] The invention belongs to the field of electronic signal processing, and relates to a human body perception method based on CSAR three-dimensional imaging of an outer ring scanning array. Background technique [0002] With the development of science and technology and the improvement of quality of life, the application scenarios of non-contact posture sensing technology are becoming more and more extensive. For example, real-time posture recognition of the elderly with inconvenient mobility in nursing homes. When an elderly person falls, the sensing system automatically Identify and judge and issue alarms to remind staff to deal with it; control home appliances and furniture by recognizing human movements or gestures in smart home scenarios; strengthen personnel management by recognizing prisoners' postures in smart prison application scenarios. [0003] At present, the mainstream method for human perception is based on the optical image captured by the...

AI Technical Summary

Problems solved by technology

[0003] At present, the mainstream method for human body perception is based on the combination of optical images captured by optical cameras and image processing algorithms for perception. This method is relatively mature, but the perception effect is greatly affected by environmental factors. The effect is low under weather conditions, and the camera captures a true reflection of real-life scenes, and there is a problem of privacy leakage

The other is a posture recognition method based on Wi-Fi wireless perception. This method mainly uses CSI amplitude information changes for detection, requires a large amount of prior data for training, and its carrier frequency and bandwidth are low, resulting in low resolution of distance information. , the recognition accuracy is low

Images