Automated Detection of Animals in Low-Resolution Airborne Thermal Imagery

Abstract

Detecting animals to estimate abundance can be difficult, particularly when the habitat is dense or the target animals are fossorial. The recent surge in the use of thermal imagers in ecology and their use in animal detections can increase the accuracy of population estimates and improve the subsequent implementation of management programs. However, the use of thermal imagers results in many hours of captured flight videos which require manual review for confirmation of species detection and identification. Therefore, the perceived cost and efficiency trade-off often restricts the use of these systems. Additionally, for many off-the-shelf systems, the exported imagery can be quite low resolution (<9 Hz), increasing the difficulty of using automated detections algorithms to streamline the review process. This paper presents an animal species detection system that utilises the cost-effectiveness of these lower resolution thermal imagers while harnessing the power of transfer learning and an enhanced small object detection algorithm. We have proposed a distant object detection algorithm named Distant-YOLO (D-YOLO) that utilises YOLO (You Only Look Once) and improves its training and structure for the automated detection of target objects in thermal imagery. We trained our system on thermal imaging data of rabbits, their active warrens, feral pigs, and kangaroos collected by thermal imaging researchers in New South Wales and Western Australia. This work will enhance the visual analysis of animal species while performing well on low, medium and high-resolution thermal imagery.

Object detection in order to determine locations for wildlife crossings

“…Manual detection relies on observers, and observers can be subject to biases and other factors, including fatigue, interest, skill level, training, eyesight etc. (Ulhaq et al, 2021). Fortunately, with the proper use of technology and software engineering, those hours of footage do not even have to be viewed in their entirety, but different deep learning methods can be used to identify objects in videos, where it is very easy to get information at what location, at what time, how much, and which animals were recorded.…”

Mask YOLOv7-Based Drone Vision System for Automated Cattle Detection and Counting

“…The recent advancement in machine learning has greatly increased the application of drone vision systems in animal detection and counting (Eikelboom et al, 2019). UAV applications cut across different tasks such as estimation of livestock population (Chabot et al, 2018;Eikelboom et al, 2019;Han et al, 2019;O'Leary et al, 2020;Ulhaq et al, 2021). Image segmentation is one of the most employed techniques for automatic detection and counting of animals in images; it works on either the instance or the semantic of objects (animals) in the images using their pixels with a specific threshold (Chabot et al, 2018;Dujon et al, 2021).…”

Dual-stream Convolutional Neural Networks for Koala Detection and Tracking

“…Similarly, Yolo series [17,16,4] splits images into grid cells and make each cell to detect objects. Yolov3 has been modified and applied to animal detection using thermal images [24], demonstrating good performance in detecting small objects.…”