Drone Swarm and Gesture Control
Utilizing DJI Ryze Tello drones, I've concentrated on developing advanced Drone Swarm Capabilities. This project enables the control of multiple drones simultaneously, either remotely or autonomously. Leveraging AI, the drone swarm can respond to hand gesture commands and autonomously follow a person's face. While the system is still developing a robust automatic collision detection feature, its current performance is impressive.
This project was also extended to demonstrate how drone swarms can be utilized in controlled environments for specific tasks, such as searching for and detecting casualty soldiers. For this purpose, I trained and developed a simplistic AI model capable of recognizing images of soldiers. This prototype marks the beginning of exploring vast possibilities in the field of autonomous drone technology and AI, showcasing potential applications in various sectors, including search and rescue operations.
More details and the overall architecture of this project can be found on Github: https://github.com/antrikshv/Tello/tree/master/Swarm/PythonTello [under autonomous swarm]
Search n Rescue
Using all the develop capabilities such as AI object detection for the drone, simplistic visual collision detection and the swarm capabilities of the drone, here is a video of an implemented controlled Search and Rescue mission where the drones are out to find soldier action figures located near the green military boxes!
AI Face Tracking
Using a simple HAAR cascading algorithm as well as tensorflow. The drone identifies a face and begins to move accordingly, whilst following where the face moves within the space
Swarm Digital Twin
USing a previous projects digital twin capabilities, implemented, multi drone twin, allowing for swarm control of drones whilst still maintaining a simulation copy giving an estimation of the drone in the virtual space.