SEE26-AI-01 AI analysis system for bee disease detection PFE

Project overview

• Develop an intelligent system based on artificial intelligence (AI) and image processing for automatic detection of bee diseases.
• Implement real-time monitoring of the hive via a mobile application, integrating visual and environmental data collected from beehives.
• Reference: SEE26-AI-01.

Objectives

• Gather and analyse visual (images/video) and environmental (temperature, humidity, etc.) data from beehives to enable early disease detection.
• Use cutting-edge computer vision, machine learning, and IoT technologies to provide accurate, robust detection and continuous monitoring.

Main tasks and scope

• Design and implement an image acquisition and preprocessing pipeline (camera calibration, denoising, segmentation of bees/frames).
• Collect, annotate and curate a dataset of beehive images/videos with disease labels; define annotation protocol and quality checks.
• Research and train ML/CV models (classification, object detection, segmentation) for detecting signs of common bee diseases.
• Integrate environmental sensor data (IoT) with visual features to improve detection accuracy and produce contextual alerts.
• Develop a mobile application prototype for real-time monitoring and alerts; implement inference pipeline for on-device or edge/cloud deployment.
• Evaluate model performance (precision, recall, ROC, latency) and iterate on optimizations for accuracy and efficiency.

Required skills and technologies

• Strong programming skills in Python; experience with OpenCV for image processing.
• Experience with deep learning frameworks (TensorFlow or PyTorch) and model training/validation workflows.
• Knowledge of IoT sensor integration, data collection from embedded devices, and basic electronics/sensor calibration.
• Familiarity with mobile app development or mobile-backend integration (native or cross-platform frameworks) and deploying ML models to mobile/edge.

Deliverables

• Curated and annotated dataset of beehive images/videos with accompanying environmental data.
• Trained and validated ML/CV models for bee disease detection and a report on model evaluation metrics.
• Prototype mobile application demonstrating real-time monitoring and alerting, plus integration plan for IoT sensors.
• Documentation including setup, deployment instructions, user guide, and recommendations for future improvements.

Application

• To apply, use the official application link: https://lnkd.in/g3V53uMH
• When applying, mention reference SEE26-AI-01 and include prior experience with computer vision, ML models, and any IoT/mobile projects.