A Hybrid RFID and CV System for Item-Level Localization of Stationary Objects

Everton Berz, Deivid Tesch, Fabiano Hessel
PUCRS University


Systems need to know the physical locations of objects and people to optimize user experience and solve logistical and security issues. Indoor positioning systems (IPSs) based on more than one technology can improve localization performance by leveraging the advantages of different technologies. This work proposes a hybrid IPS able to estimate the item-level location of stationary objects using off-the-shelf equipment. By using RFID technology, machine learning approaches based on artificial neural networks (ANNs) and support vector regression (SVR) are proposed. A k-means technique is also applied to improve accuracy. A computer vision (CV) subsystem detects visual markers in the scenario to enhance RFID localization. To combine the RFID and CV subsystems, a fusion method based on region of interest (ROI) is proposed. We have implemented our system and evaluated it using real experiments. The localization error is between 9 and 29 cm in the range of 1 and 2.2 m scenarios. In a machine learning approach comparison, ANN performed 31% better than SVR approach.