Akademik Veri Yönetim Sistemi
2024 American Society of Agricultural and Biological Engineers Annual International Meeting, ASABE 2024, California, Amerika Birleşik Devletleri, 28 - 31 Temmuz 2024, (Tam Metin Bildiri)
Unmanned Aerial Vehicles (UAVs) play a crucial role in a myriad of agricultural and infrastructural applications, encompassing surveillance, observation, and spraying, etc.. Nevertheless, challenges persist in the realm of limited payload distribution and insufficient operation time, particularly when dealing with large acres of farmland and multiple tasks. The utilization of multiple UAVs in a swarm configuration allows for distributed payload and task capabilities, thereby extending the overall flight duration. However, UAVs forming a swarm, posing issues related to formation and localization, particularly the presence of GPS denied zones. Leader-follower drone system presents a formidable obstacle, necessitating innovative solutions for formation control to address such localization challenges. In addressing these challenges, our research introduces a vision-based relative navigation algorithm for leader-follower formation control. This approach employs machine vision methods for localization with blob operation and machine learning object detection, enabling the follower UAV to seamlessly track the leader drone. Furthermore, the follower UAV can transmit its relative position to the leader UAV, facilitating mutual localization. To validate the efficacy of our proposed method, we conducted experiments wherein inspection UAVs were modeled and implemented in Webots environments. Specifically, the leader UAV was subjected to a GPS denied zone, prompting the follower UAV to send relative position data to aid in re-localization and restore the desired position. This innovative vision-based approach holds promise for enhancing the autonomy and localization capabilities of UAV inspection in challenging operational environments.