Developing CAMIC-Behav:A Computer Vision Framework for Behavioral Monitoring of Dromedary Camels
Alhashmi, Fawaghy Ahmed Saeed Mohamed
Alhashmi, Fawaghy Ahmed Saeed Mohamed
Supervisor
Department
Computer Vision
Embargo End Date
2025-05-30
Type
Thesis
Date
2025
License
Language
English
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
In the UAE, dromedary camels are of significant cultural and economic importance, yet monitoring their behavior remains a challenging task due to the limited specificity of existing automated systems. This thesis introduces CAMIC-Behav, a novel dataset created to enhance the automated behavioral analysis of dromedary camels using a deep learning framework that integrates the YOLO and SAMURAI models. Our methodology involves the collection of extensive behavioral data under varying environmental conditions, followed by the application of YOLOv11 for behavior detection and SAMURAI for tracking. The system demonstrated a mean average precision (mAP) of 80.9% with the ability to perform realtime monitoring. These advancements significantly enhance the accuracy and efficiency of behavior analysis over existing methods, offering a powerful solution for real time camel behavior monitoring. This framework can also be adapted for other livestock, potentially improving animal welfare.
Citation
Fawaghy Ahmed Saeed Mohamed Alhashmi, “Developing CAMIC-Behav:A Computer Vision Framework for Behavioral Monitoring of Dromedary Camels,” Master of Science thesis, Computer Vision, MBZUAI, 2025.
Source
Conference
Keywords
Camel Monitoring with Intelligent Cameras (CAMIC), AI Artificial Intelligence, CNN Convolutional Neural Network., LSTM Long Short-Term Memory., SAMURAI Segment Anything Model for Zero-shot Visual Tracking with Motion- Aware Memory.