Network visibility graph-based human body orientation detection method

Abstract

The invention relates to a network visibility graph-based human body orientation detection method. According to the method, ordinary equipment is adopted to build an indoor wireless channel information data collection platform; positioning is divided into two phases, namely, an offline training phase and an online testing phase. According to the offline phase, the channel state information data ofa human body in different orientations are collected; the channel state information data are preprocessed; networking processing is performed on the data; network features are extracted; standardization processing is performed on the network features, and the processed network features are stored in a fingerprint database of different orientations; and an orientation-data fingerprint mapping relationship is built. According to the online phase, the data are processed in the same manner as the above mentioned, and a machine learning algorithm is used to perform classification detection on testdata; and in order to test the accuracy of classification, the amplitude and phase information of the channel state information data are comprehensively utilized to observe classification results under three cases. With the method of the invention adopted, the orientation detection of people in an indoor environment can be realized at a low cost. The method of the invention has a certain application value in the fields such as somatosensory games and the orientation identification of blind people.

Description

technical field [0001] The invention relates to the field of human body perception, in particular to a method for detecting human body orientation based on network visual graphs. Background technique [0002] Orientation detection is an important aspect of human perception. If the specific orientation of the human body can be judged and estimated, the current activity status of the person can be deduced. [0003] The detection of passive human orientation has good application prospects in many fields. In terms of somatosensory games, most of the sensing technologies currently rely on specific sensors. The cumbersome equipment wearing process and relatively expensive equipment costs have had a certain impact on the popularity of somatosensory games. The use of wireless signals to judge human body status can improve player In terms of business intelligence, how to determine the consumption point of interest of customers in a large shopping mall with a huge flow of people is o...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the disadvantages of complex deployment, weak anti-interference ability, and high price of the existing human body orientation detection methods, the present invention proposes a human body orientation detection method based on network visual graphs, which realizes the detection of human body orientation and is simple to deploy , strong anti-interference ability, and low price

Images