A Low-Cost Non-Invasive RFID-Based Framework for Human Posture Detection

This paper presents an RFID-based system for identifying various human postures using a few passive RFID tags integrated with furniture, such as a chair. Posture detection can be achieved through a simple and straightforward algorithm based on the error calculation between test data set and a reference one, encoded by the states of the tags. These states are represented by the normalized counting of tag EPC by a reader for a specific time duration. The algorithm selects the correct posture by identifying the test pattern with minimum error from the reference. The placement and quantity of RFID tags are optimized to balance system accuracy and complexity. The proposed method can reliably distinguish between different sitting postures achieving an accuracy of up to 90 - 94%. Also, a possible framework of extending this method to identify sleep posture in similar way.