Deep Learning Enhanced Hand Gesture Recognition for Efficient Drone use in Agriculture

The use of deep learning in unmanned aerial vehicles (UAVs), or drones, has greatly improved various technologies by making complex tasks easier, faster, and requiring less human help. This study looks into how artificial intelligence (AI) can be used in farming, especially through creating a system where drones can be controlled by hand gestures to support agricultural activities. By using a special type of AI called a Convolutional Neural Network (CNN) with an EfficientNet B3 model, this research developed a gesture recognition system. It was trained on 1,393 pictures of different hand signals taken under various light conditions and from three different people. The system was evaluated based on its training and testing performance, showing very high scores in terms of loss, accuracy, F1 score, and the Area Under the Curve (AUC), which means it can recognize gestures accurately and work well in different situations. This has big implications for farming, as it gives farmers an easy way to control drones for tasks like checking on crops and spraying them precisely, which also helps keep them safe. This study is an important step towards smarter farming practices. Moreover, the system’s ability to perform well in different settings shows it could also be useful in other areas like construction, where drones need to operate precisely and flexibly.

@article{Srinil2024,
  title     = {Deep Learning Enhanced Hand Gesture Recognition for Efficient Drone use in Agriculture},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {15},
  number    = {5},
  year      = {2024},
  publisher = {The Science and Information Organization},
  author    = {Phaitoon Srinil and Pattharaporn Thongnim},
  doi       = {10.14569/IJACSA.2024.01505127},
  url       = {https://doi.org/10.14569/IJACSA.2024.01505127}
}