System and method for recognizing and monitoring exercising action through unbound radio frequency label

Abstract

The invention provides a system and a method for recognizing and monitoring exercising action through an unbound radio frequency label. The system comprises a plurality of commercial passive RFID label, an antenna, a Reader, a CPU and a prompting unit; the CPU comprises a data processing unit and a data storing unit; the detected people or matter within the communication range of the label and the reader can be recognized without modifying any existing commercial device or extra hardware facilities; the method is simple to use, quick and effective, can monitor and supervise the exercising action of a user on real time and guide the user. With the adoption of the method, the action standardability can be recognized, and the exercising action of a human body can be recorded and monitored on real time; the standard action data of the user are acquired firstly; the feature wave configuration files are extracted through a plurality of filters and by the extraction algorithm; then the configuration files can be matched on real time through the DTW algorithm to recognize the user action when the user exercises; the test performed in an indoor and closed office building, the recognizing precision of the exercising action is up to 92%.

Description

technical field [0001] The invention relates to the field of sports training auxiliary equipment, in particular to a phase difference-based non-binding radio frequency tag fitness action recognition and monitoring system. Background technique [0002] At present, there are many researches based on motion recognition. In RFID, such as the AllSee system developed by the University of Washington, the RSSI envelope is extracted from the RF signal to recognize 8 pre-defined simple gestures with high precision, but it modifies the hardware equipment, and It can only recognize simple gestures, and because AllSee considers low energy consumption, it is easily interfered by other factors, resulting in inaccurate recognition results; the Tagoram system developed by Tsinghua University uses phase difference and hologram technology to track the movement of objects in real time, making positioning It has extremely high accuracy in tracking and tracking, but it belongs to binding position...

AI Technical Summary

Problems solved by technology

[0002] At present, there are many researches based on motion recognition. In RFID, such as the AllSee system developed by the University of Washington, the RSSI envelope is extracted from the RF signal to recognize 8 pre-defined simple gestures with high precision, but it modifies the hardware equipment, and It can only recognize simple gestures, and because AllSee considers low energy consumption, it is easily interfered by other factors, resulting in inaccurate recognition results; the Tagoram system developed by Tsinghua University uses phase difference and hologram technology to track the movement of objects in real time, making positioning It has extremely high accuracy in tracking and tracking, but it belongs to binding positioning, which requires the user's body to be labeled, and uses multiple antennas arranged in different directions, so the application scenarios are very limited

In the field of wifi, the WiTrack and WiSee systems developed by the Massachusetts Institute of Technology in the United States can recognize gestures or movements of the human body through walls. WiSee uses FFT and OFDM technology to extract Doppler shift information from The change of Le is mapped to the difference of gestures, which can recognize 9 gestures; WiTrack transforms USRP equipment to recognize human body movement and recognize simple gestures at the same time, but they all need to use expensive USRP, and require multiple antennas at the same time. Requires modification of existing equipment

[0003] The above is the most popular and cutting-edge gesture recognition technology at present. Some other gesture recognition technologies, such as image-based processing technology and infrared remote sensing technology, are also very mature, but they cannot be applied to the RFID field.

Images