Camera-based Virtual Drone Control System Using Two-handed Gestures

Abstract

As Industry 4.0 has grown, which aims to improve efficiency and productivity by optimizing operations using advanced technologies such as artificial intelligence (AI), the Internet of Things (IoT), and robotics, the use of drones and autonomous vehicles as innovative transportation methods is increasing. However, autonomous vehicles encounter numerous unforeseen events when driving on the ground with many obstacles, and additional transportation means are required when transporting vehicles in remote areas such as islands. On the other hand, drones face fewer obstacles in the sky, there are no cases where they cannot enter due to the characteristics of the terrain or require additional transportation. But, there are some problems. Firstly, they are not yet commercialized although research on autonomous drones is advancing. Secondly, controlling drones proficiently using a remote controller requires skilled operators, leading to accidents caused by human error. Lastly, drones cannot be controlled without a remote controller. Therefore, we propose a camera-based virtual drone control system based on hand gestures. To achieve this, we defined gestures and it was demonstrated through virtual drone control in a virtual environment.