基於RFID與視覺技術之出入口人員追蹤識別系統

People Tracking and Identification System Based on RFID and Visual Techniques at the Entrance

陳又銘(Yu-Ming Chen)

指導教授 ： 廖珗洲

朝陽科技大學/資訊學院/資訊工程系/碩士(2013年)

摘要

近年來，無線射頻辨識技術(RFID: Radio Frequency Identification)的應用已經非常普遍了，也有許多研究將RFID與視覺技術做結合，來擴展RFID技術的應用領域。本研究是針對配戴主動式RFID Tag通過出入口的情境下，提出一套方法來運用Tag的訊號強度(Received Signal Strength Indication, RSSI)，可以結合攝影機的即時影像，達成物體的鎖定識別與追蹤的目標。
此研究中主要挑戰是RSSI訊號是非常不穩定的，而且不同的Tag其RSSI資訊有很大的差異。因此，方法的設計上首先將RSSI資訊利用卡爾曼濾波器來處理，讓訊號較為穩定，接著設計一個適應性的門檻值來決定是否要進行影像上物體的鎖定，鎖定過程中也依據Tag通過出入口的變化特性，設計了一個訊號衰減檢測機制來確認鎖定的目標是否有誤，發現鎖定錯誤的時候會重新進行鎖定的動作。最後，針對實作的雛形系統所做的實驗結果顯示，本研究所提出方法可以有效地克服RSSI訊號不穩定的問題，達成出入口人員進出鎖定識別與追蹤的目標。

關鍵字

卡爾曼濾波器 ； 視覺追蹤 ； 室內定位 ； RFID

並列摘要

Recently, the RFID (Radio Frequency Identification) technology is widely used in many areas. Some studies combine the RFID and visual techniques to extend the application domains. In this study, a method is proposed to locate, identify, and track persons with active RFID tag while they passing through an entrance.
The main challenge of this study is the unstable RSSI (Received Signal Strength Indication) values of an RFID tag. The RSSI information is also significant difference among tags. Firstly, a Kalman filter is used to smooth the received RSSI values. Then, an adaptive threshold of RSSI is designed. When the RSSI is larger than the threshold, it triggers the locating of the target person on the image. However, the located target may be wrong since the locating process is triggered too early and another person is coincidently passing the entrance. Therefore, an RSSI attenuation checking mechanism is designed based on the temporal features of the RSSI values while a person passing the entrance. The locating process will be restarted if the checking cannot pass. An experimental study is also designed to evaluate the performance of the proposed method. The results show that the implemented prototype based on the method can overcome the problem caused by the unstable RSSI values and achieve the purpose of this study to locate, identify, and track persons while they passing through the entrance.

並列關鍵字

Kalman Filter ； Indoor Positioning ； Visual Tracking ； RFID

參考文獻

[1] A. Sani, M. Rajab, R. Foster and Y. Hao, “Antennas and Propagation of Implanted RFIDs for Pervasive Healthcare Applications,” Proceedings of the IEEE, vol. 98, no. 9, 2010, pp. 1648-1655.

[2] G.B. Gentili, F. Dori and E. Iadanza, “Dual-frequency active RFID solution for tracking patients in a children''s hospital. Design method, test procedure, risk analysis, and technical solution,” Proceedings of the IEEE, vol. 98, no. 9, 2010, pp. 1656-1662.

[3] C. S. Wang and L. C. Cheng, “RFID & vision based indoor positioning and identification system,” Proceedings of the 2011 IEEL 3rd International Conference on Communication Software and Networks (ICCSN), Xi''an, China, 27-29 May 2011, pp. 506 – 510.

[4] C. S. Wang, C. H. Huang, Y. S. Chen and L. J. Zheng, “An implementation of positioning system in indoor environment based on active RFID,” Proceedings of the 2009 Joint Conferences on Pervasive Computing (JCPC), Tamsui, Taipei, 3-5 Dec. 2009, pp. 71-76.

[5] S. H. Cheng, “An indoor positioning system based on active RFID in conjunction with Bayesian network,” Proceedings of the 2011 International Conference on Machine Learning and Cybernetics (ICMLC), Guilin, China , 10-13 July 2011, vol. 1, pp. 386-390.