Object Tracking and Following with a Drone Using Pixhawk and ROS2 w/ XRCE-DDS - Farhang Naderi, URI
2.3 هزار بار بازدید -
8 ماه پیش
-
In this enlightening session, I
In this enlightening session, I will present process of constructing a sophisticated drone control system, which incorporates a Pixhawk flight controller, Raspberry Pi CM4, and the usage of ROS2. The system's standout capability is its ability to track and follow objects using a camera connected to the CM4. Image processing and object detection are handled by the CM4, utilizing machine learning algorithms with a particular emphasis on applying convolutional neural networks (CNNs). This continuous processing of the camera feed allows the drone to react to the movements of the tracked object dynamically. A significant technical challenge addressed in this presentation is the direct communication between ROS2 and Pixhawk for control commands, which forms a crucial part of the control logic. This interface allows the drone to adjust its trajectory in real-time based on the object's detected movement. All the decisions coming from the Machine Learning algorithm is fed to PX4 as attitude controlling messages such that the drone can follow the object accordingly.
8 ماه پیش
در تاریخ 1402/10/01 منتشر شده
است.
2,337
بـار بازدید شده