Tracking And Localization Of Moving Objects With UHF RFID For Service Robots

Service robots have high potential of assisting the work places in different areas like shopping malls,libraries,hospitals,factories,and office environments.Service robots have to localize and track moving objects to perform assistance and guidance.Localization of moving objects like humans can be a complex and challenging task as the movements can be very rapid and unpredictable.Recent articles show that RFID(Radio Frequency IDentification)technology is rapidly emerging for the localization of moving objects.Fusion of RFID technology and laser makes them particularly suitable for localization of objects in industrial environments(e.g.asset management).In this thesis,we address the issue of tracking and localization of objects with two different sensory information namely RFID and laser range finder.We have introduced two novel methods to address localization of moving objects(moving humans)and conducted extensive set of experiments to validate proposed methods.These experiments have been conducted by using a service robot named as SCITOS G5 which has different on-board sensors including RFID readers,camera,laser range finder etc.First method is based on fusion of RFID phase and laser clustering.We have computed the RFID phase-based velocity by the phase difference.Laser data is clustered and then laserbased velocity is calculated.Afterwards,a particle filter is used for the sensor fusion based on the best K matching clusters.We conducted extensive experiments to validate the feasibility of this method and our results show that a mean positioning accuracy up to 0.25 m can be achieved through this method.Second method is based on an RFID tag array and laser ranging information.We have observed in first method that radio signals can be blocked by human body at some positions so the localization accuracy achieved with a single tag is limited.Specifically,radio signal blockage is occurred in an environment where multiple humans are moving in the test area.So in this method,we encircled a human with a tag array and calculated the phase-based radial velocity of all tags.In this method,DBSCAN(Density-Based Spatial Clustering of Applications with Noise)algorithm is used for laser clustering and then laser-based radial velocity is calculated.Laser-based radial velocity is matched with phase-based radial velocity to get best matching clusters.A particle filter is used to localize the moving human by fusing the matching results of both velocities.The results verified the feasibility of this method and proved that our approach significantly increases localization accuracy by up to25% compared to a single tag approach.These new and innovative localization methods expressed in this thesis can be very useful for different industries like shopping malls,hospital environment,library,offices and factories.These methods can be implemented on a broad line of applications to use service robots and RFID technology to satisfy real time industry needs.