UAV-Based IoT Applications for Action Recognition

The Internet of Things (IoT) concept permits the connection of any device using any network to deliver services. The rapid development of IoT has capitalized the integration of new devices such as Drones which gives bird’s eye view. In fact, Drones or Unmanned aerial vehicles (UAV) have seen a tremendous increase in the last few years. It is possible to use UAVs for a variety of tasks, including surveillance, delivery, and search and rescue. It is important in these applications to analyze human behavior from an aerial point of view. Therefore, the number of approaches for analyzing actions from aerial videos has emerged. A number of datasets have been proposed for the development and the evaluation of these approaches. Nevertheless, currently there is no survey devoted to the subject of human action recognition from aerial videos. Hence, we tried to cover the lack of a complete description of the most important aerial dataset and present an overview of the state-of-the-art approaches proposed on this field. The detailed analysis of these datasets and approaches is an essential reference for future research.