Font Size: a A A

Device-free Target Monitoring System Using RFID Signatures

Posted on:2021-02-20Degree:MasterType:Thesis
Country:ChinaCandidate:B GuFull Text:PDF
GTID:2428330614966049Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the continuous development of Internet of things(iot)technologies with typical applications such as Internet of vehicles,smart home and smart medical care,efficient scene perception has attracted extensive attention of researchers.Scene sensing monitoring systems can be divided into device-based monitoring and device-free monitoring.Among them,device-based monitoring requires the tested target to carry wearable devices at all times,while device-free monitoring mainly uses wifi or Radio Frequnency Identification(RFID)for scene monitoring,without requiring users to carry any communication equipment,showing a good application prospect.This thesis mainly studies the contactless scene monitoring technology based on RFID.In view of the different application of RFID,this thesis presents two monitoring systems based on RFID,namely liquid level monitoring system and detection of hand approaching monitoring system.In rfidbased liquid level monitoring system,the system uses the collected signal strength as the criterion for determining the liquid level status.The threshold corresponding to the liquid level status is obtained in the training stage,and then the liquid level in the measuring tank is monitored in the real-time monitoring stage.In view of the weak anti-interference capability of RSSI,the Local Weight linear Regression(LWR)algorithm is introduced to extract more stable RSSI features.In the rfid-based detection of hand approaching monitoring system,the attention of goods is analyzed by monitoring the hand activities of people in the environment.The system is deployed in a single object and multiobject scenario,combining an adaptive segmentation alogorithm based on kullback-leibler(KL)to achieve action recognition,and combining the coefficient of variation to achieve target recognition.The experimental results show that the corresponding accuracy of the two monitoring systems constructed in this thesis is 94.76% and 96% respectively,which have higher detection accuracy.
Keywords/Search Tags:Device-free monitoring, RFID, Liquid Level Monitoring, Detection of Hand Approaching monitoring, LWR, KL
PDF Full Text Request
Related items