Improved locating algorithm of railway tunnel personnel based on collaborative information fusion in Internet of Things

With larger-scale of railway construction in China, there are more and more larger-scale tunnel projects in process. Tunnel engineering constructions are very difficult, high risk, and there are many unpredictable factors that may cause safety issues, such as landslides, roof caving and water bursts, threatening the construction personnel's safety. The personnel positioning tracking systems have been studied and applied preliminarily in railway tunnel construction. It is important to know how many people are underground, and workers must sign in/out one by one at tunnel entrances when they come in or leave; a process that is time-consuming and sometimes irritating to those who line up and must exit one-by-one to sign in. Active Radio Frequency Identification (RFID) systems can respond to transponders on the personnel's helmets or jackets to quickly identify workers coming and going without stopping to sign in and out. It can also alert management when a person enters without a transponder. Indoor moving target recognition and tracking in Internet of Things is a popular research and application topic in recent years. This paper proposes an indoor moving target recognition and tracking method based on RFID and Charge Coupled Device (CCD) collaborative information fusion. First, RFID technique and the proposed extended virtual reference elimination (extended virtual reference elimination) approach are used to recognize and to coarsely locate the target. Second, based on the coarse localization results, the monitoring/sleeping control of different CCDs will be realized. Subsequently, the background-difference method is used to detect the target in CCD monitoring image and realize precise localization with multiple angle of view fusion. Finally, with weighted average of the two localization results, the moving target location is obtained. The method combines the advantages of RFID fast recognition and localization and CCD precise localization. The experiment results indicate that the proposed collaborative information fusion method can effectively improve the accuracy and real-time performance of indoor moving target tracking.