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.


    Access

    Download


    Export, share and cite



    Title :

    Advanced Computer Vision Methods for Tracking Wild Birds from Drone Footage


    Contributors:


    Publication date :

    2024




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    Unknown




    Method of remotely obtaining drone footage.

    HANSCOM ERIC | European Patent Office | 2020

    Free access

    Method of remotely obtaining drone footage

    HANSCOM ERIC | European Patent Office | 2021

    Free access

    Advanced computer vision for extracting georeferenced vehicle trajectories from drone imagery

    Fonod, Robert / Cho, Haechan / Yeo, Hwasoo et al. | ArXiv | 2024

    Free access

    Drone for Developing Vision Tracking and Tracking Module

    PARK JONG DEOK | European Patent Office | 2021

    Free access

    Air drone device for tracking and observing wild animal

    CHOI JIN HO | European Patent Office | 2022

    Free access