Individual recognition of Eurasian beavers (Castor fiber) by their tail patterns using a computer-assisted...

Area of Science:

• Ecological sciences
• Wildlife biology
• Computer vision applications in ecology

Background:

• Individual recognition is crucial in ecological studies.
• Photograph-based methods offer non-invasive animal identification.
• Advancements in computer vision enable faster analysis of large image datasets.

Purpose of the Study:

• To evaluate the effectiveness of the Scale Invariant Feature Transform (SIFT) algorithm for individual beaver recognition.
• To test a semi-automated method for analyzing tail pattern images.
• To establish a reliable, non-invasive technique for monitoring Eurasian beaver populations.

Main Methods:

• Applied the SIFT algorithm to 800 tail scale pattern images from 100 Eurasian beavers (Castor fiber).
• Extracted scale patterns from dorsal tail images using open-source image processing software.
• Utilized an 80% training and 20% testing data split for algorithm validation.

Main Results:

• Achieved an overall recognition accuracy of 95.7%.
• Demonstrated that individual scale patterns on the beaver's tail are unique and identifiable.
• Confirmed the robustness of the SIFT algorithm to variations in image acquisition.

Conclusions:

• The SIFT algorithm is a highly effective tool for distinguishing individual beavers based on tail scale patterns.
• This method provides a scalable and non-invasive approach for wildlife monitoring and ecological research.
• Photograph-based analysis using SIFT can significantly enhance the efficiency of large-scale animal identification studies.