Personnel action recognition and position estimation method based on channel state information

Abstract

The invention discloses a personnel action recognition and position estimation method based on channel state information. The method comprises steps: in the offline stage, carrying out different actions on different reference position points by utilizing a target, and receiving CSI of a WiFi signal; constructing a CSI image according to the time domain, spatial domain and frequency domain information of the CSI measurement value amplitude, and forming an action recognition training data set and a position estimation training data set associated with the CSI image; and finally, performing action-based classification learning by using a convolutional neural network CNN to obtain an action classification model, extracting depth feature information at two axis positions of a CSI image by usinga CNN, and performing regression learning by using a support vector machine SVM to obtain regression functions based on the two axis positions respectively; in an online stage, constructing a CSI image by using the received CSI measurement value, and estimating a target action by using an action classification model; and meanwhile, extracting depth feature information on two-axis positions of theCSI image, and calculating a target position by utilizing a regression function. The method is high in positioning precision and action recognition rate, simple in structure and low in cost.

Description

technical field [0001] The invention relates to a positioning recognition technology, in particular to a method for realizing target action recognition and position estimation under the condition of no equipment by using channel state information of Wifi signals and machine learning. Background technique [0002] The location and state estimation of persons has various applications in daily life, such as elderly monitoring, person tracking, early warning in wards, etc., and has attracted extensive attention from academia and industry. The traditional realization of personnel state and position estimation either adopts the method of wearable devices or relies on cameras. However, both methods have certain application limitations. Most of the methods that require wearable devices require the use of devices with built-in wireless sensors. These devices are not only expensive, but also inconvenient to wear in some occasions; while the methods that rely on cameras need to rely o...

AI Technical Summary

Problems solved by technology

However, both methods have certain application limitations.

Most of the methods that require wearable devices require the use of devices with built-in wireless sensors. These devices are not only expensive, but also inconvenient to wear on some occasions; while the methods that rely on cameras need to rely on conditions such as lighting, and there are dead spots in the camera

Images