Bearded Dragon Activity Recognition Pipeline: An AI-Based Approach to Behavioural Monitoring

2025-07-23

AI Summary

To address the time-consuming, subjective, and low-precision nature of manual observation in bearded dragon behavioral monitoring, this study proposes a real-time video analytics system based on YOLOv8s. The framework innovatively integrates object detection, frame-level coordinate temporal interpolation, and a rule-based logic engine to enable automated recognition of basking and foraging behaviors. Compared with conventional approaches, it significantly improves behavioral discrimination continuity and robustness. Experimental results show an mAP@0.5:0.95 of 0.855 for behavior-level detection. Basking recognition is stable and reliable, whereas foraging detection performance is currently limited by cricket detection accuracy (mAP@0.5 = 0.392), highlighting a clear direction for future improvement—namely, multi-object fine-grained detection. This work establishes a scalable technical paradigm for quantitative, fine-grained behavioral analysis of reptiles.

Technology Category

• Computer Vision: Video Understanding & Activity Analysis
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Web Mining and Content Analysis: Topic discovery and tracking

Abstract

Traditional monitoring of bearded dragon (Pogona Viticeps) behaviour is time-consuming and prone to errors. This project introduces an automated system for real-time video analysis, using You Only Look Once (YOLO) object detection models to identify two key behaviours: basking and hunting. We trained five YOLO variants (v5, v7, v8, v11, v12) on a custom, publicly available dataset of 1200 images, encompassing bearded dragons (600), heating lamps (500), and crickets (100). YOLOv8s was selected as the optimal model due to its superior balance of accuracy (mAP@0.5:0.95 = 0.855) and speed. The system processes video footage by extracting per-frame object coordinates, applying temporal interpolation for continuity, and using rule-based logic to classify specific behaviours. Basking detection proved reliable. However, hunting detection was less accurate, primarily due to weak cricket detection (mAP@0.5 = 0.392). Future improvements will focus on enhancing cricket detection through expanded datasets or specialised small-object detectors. This automated system offers a scalable solution for monitoring reptile behaviour in controlled environments, significantly improving research efficiency and data quality.

Problem

• Automate bearded dragon behavior monitoring using AI
• Improve accuracy of basking and hunting detection
• Enhance cricket detection for better hunting analysis

Innovation

• Uses YOLO models for real-time video analysis
• Trains on custom dataset with 1200 images
• Applies rule-based logic for behavior classification