Gautham Manoharan
Real-Time Object Detection for Autonomous Drone Navigation and Obstacle Avoidance
Programmed DJI Tello to classify objects in vision, identify obstacles and interpret hand pose to navigate
Work
- Leveraged OpenCV, Mediapipe, Tensorflow, and YOLO for object identification and human body pose estimation.
- The Python Mediapipe library is a user friendly framework for computer vision.
- Leveraging Mediapipe, a model to detect hand pose was built and the inputs were used to control the drone.
Videos and Images
Gesture Detection and Face Detection:
Drone control based on Hand Pose:
Object classification with YOLO:
