Queuing Recognition using Smart Shopping Carts

Recently, queuing recognition becomes an emerging research topic. Its main purpose is to recognize human's queuing behaviors using sensor devices. However, previous studies take smartphones as the major sensor devices, causing it to suffer from inefficiency and inconvenience problems in supermarket scenarios. In view of this, this work takes account of queuing behaviors in supermarkets and proposes a novel framework for queuing recognition based on smart shopping carts. The proposed framework includes four main modules: (1) queuing micro-activity recognition, (2) queuing state recognition, (3) queuing group recognition (4) queuing time inference. We implement the proposed framework and conduct simulations for queuing scenarios to evaluate the proposed framework. Results show that the proposed framework has good efficiency and achieves high accuracy.