Drone Applications in Wildlife Research—A Synoptic Review

Mazumdar, Subhendu

doi:10.1007/978-981-19-2083-7_14

Cited by 2 publications

(2 citation statements)

References 133 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The study and conservation of wildlife have entered a new era with the integration of advanced technologies, providing researchers with unprecedented tools to monitor and understand animal behavior in their natural habitats [1]. Among these technologies, unmanned aerial vehicles (UAVs) have emerged as versatile platforms for ecological research, offering the ability to capture high-resolution imagery and video from vantage points that were once inaccessible [2].…”

Section: Introductionmentioning

confidence: 99%

“…In the realm of ornithological research, the utilization of UAVs holds immense potential for tracking and studying bird species in their natural environments [1,4]. When UAVs are coupled with automated computer vision methods, particularly those based on deep learning architectures, researchers can automate the identification and tracking of bird species in drone-captured footage, thereby overcoming the limitations of traditional manual tracking methods and eliminating the dangers of physical monitoring or the limitations of using a telescope.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Advanced Computer Vision Methods for Tracking Wild Birds from Drone Footage

Mpouziotas,

Karvelis,

Stylios

2024

Drones

View full text Add to dashboard Cite

Wildlife conservationists have historically depended on manual methods for the identification and tracking of avian species, to monitor population dynamics and discern potential threats. Nonetheless, many of these techniques present inherent challenges and time constraints. With the advancement in computer vision techniques, automated bird detection and recognition have become possible. This study aimed to further advance the task of detecting wild birds using computer vision methods with drone footage, as well as entirely automating the process of detection and tracking. However, detecting objects from drone footage presents a significant challenge, due to the elevated altitudes, as well as the dynamic movement of both the drone and the birds. In this study, we developed and introduce a state-of-the-art model titled ORACLE (optimized rigorous advanced cutting-edge model for leveraging protection to ecosystems). ORACLE aims to facilitate robust communication across multiple models, with the goal of data retrieval, rigorously using various computer vision techniques such as object detection and multi-object tracking (MOT). The results of ORACLE’s vision models were evaluated at 91.89% mAP at 50% IoU.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%