BIO-EYE: AN ADVANCED WILDLIFE DETECTION FRAMEWORK

Authors

  • K.Ramya Laxmi Author
  • Maganti Nikhil Reddy Author
  • Patnam Akshitha Reddy Author
  • Pagadala Sandeepa Author
  • Lingamgunta Hemanth Author

DOI:

https://doi.org/10.62643/

Keywords:

Camera Traps, CNN, Wildlife Spotter,, Graph-Cut algorithm, Biodiversity, Species Detection,, Sustainable conservation.

Abstract

Efficient wildlife monitoring is essential for conservation and management, with
camera traps being a key tool for unobtrusive and continuous data collection.
However, the manual processing of vast amounts of camera trap imagery is labor-
intensive and time consuming. This paper proposes a framework utilizing deep
convolutional neural networks (CNNs) for automated species recognition in the wild,
based on a dataset from the Wildlife Spotter project. The system uses advanced graph-
cut algorithms to segment images and identifies species automatically, significantly
outperforming traditional methods like the bag of visual words model. This approach
not only accelerates research and enhances citizen science initiatives but also
improves management decisions. While there are challenges related to environmental
conditions and species diversity, the proposed method marks a substantial
advancement in automated wildlife monitoring. Integrating deep learning with camera
trap technology promises to revolutionize ecological research, offering more accurate
and efficient species identification and facilitating timely conservation actions.

Downloads

Published

15-01-2025

How to Cite

BIO-EYE: AN ADVANCED WILDLIFE DETECTION FRAMEWORK. (2025). International Journal of Engineering Research and Science & Technology, 21(1), 28-39. https://doi.org/10.62643/