Towards Automated Biometric Identification of Sea Turtles (Chelonia mydas)

TL;DR

This paper explores automated biometric identification of sea turtles using camera images and learned image descriptors, achieving 65% accuracy with Histogram of Oriented Gradients (HOG).

Contribution

It introduces a novel approach for sea turtle identification using robust image descriptors, highlighting HOG's effectiveness over keypoint-based methods.

Findings

HOG outperforms keypoint-based descriptors in classification accuracy.

High-dimensional descriptors are necessary due to minimal image information.

Achieved 65% average classification accuracy with HOG.

Abstract

Passive biometric identification enables wildlife monitoring with minimal disturbance. Using a motion-activated camera placed at an elevated position and facing downwards, we collected images of sea turtle carapace, each belonging to one of sixteen Chelonia mydas juveniles. We then learned co-variant and robust image descriptors from these images, enabling indexing and retrieval. In this work, we presented several classification results of sea turtle carapaces using the learned image descriptors. We found that a template-based descriptor, i.e., Histogram of Oriented Gradients (HOG) performed exceedingly better during classification than keypoint-based descriptors. For our dataset, a high-dimensional descriptor is a must due to the minimal gradient and color information inside the carapace images. Using HOG, we obtained an average classification accuracy of 65%.