Animal conservation applications for wildlife detection and classification using various machine learning algorithms

Tracking animal behaviour, habitat use, population demographics, conflicts between humans and animals, vehicle-animal accidents, events involving snares and poaching, and illnesses all depend on wildlife monitoring. Over time, there have been an increasing number of incidents between vehicles and an...

Full description

Saved in:
Bibliographic Details
Main Authors: Murugesan, Mangaleswaran, Mahendiran, Azhagiri
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
Animals
Artificial neural networks
Demographics
Image processing
Machine learning
Monitoring
Roads & highways
Rural roads
Wildlife conservation
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Tracking animal behaviour, habitat use, population demographics, conflicts between humans and animals, vehicle-animal accidents, events involving snares and poaching, and illnesses all depend on wildlife monitoring. Over time, there have been an increasing number of incidents between vehicles and animals on both mountainous and rural roads and highways. Occasionally, animals that cross the road at the wrong moment result in tragic collisions. The identification and categorization of animals from pictures and videos is a popular area of study. A crucial part of the methods used today for wildlife monitoring is image processing. Since most wild creatures are most active at night, it might be difficult to view them without specialist equipment. The availability of heat sensing technology and technical advancements have made it possible for researchers to observe animals. This study compares a number of methods, including deep neural network-based (Faster RCNN and YOLO) and conventional (HOG/SVM) approaches, for identifying animals in thermal camera images. Through comparison, the best mean Average Precision (mAP) was found for different Intersection over Union (IoU) coverage thresholds and recall (sensitivity) levels.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0211389