University of Twente Student Theses
As of Friday, 8 August 2025, the current Student Theses repository is no longer available for thesis uploads. A new Student Theses repository will be available starting Friday, 15 August 2025.
AI detection & counting of mosquito eggs
Syriopoulos, T. (2025) AI detection & counting of mosquito eggs.
|
PDF
39MB |
| Abstract: | Mosquito-borne diseases cause over 700,000 deaths annually according to the World Health Organisation (WHO), with effective monitoring requiring labor-intensive manual egg counting from ovitraps. This research investigates the extent to which a custom AI computer vision model trained with synthetic data on YOLOv8n architecture can automate mosquito egg detection and counting. The study developed a novel synthetic data generation pipeline with 73 configurable parameters to address the critical challenge of limited training data in specialized biological object detection. Through international collaboration across five European countries (Cyprus, Hungary, Albania, Italy, Croatia), real-world validation data was collected and annotated by entomologists from the VectorAI COST Action network. The final YOLOv8n model trained on solely synthetic data achieved 77.6% mAP50 performance with 97.7% precision, representing substantial improvement over baseline implementations for 2-8 pixel mosquito eggs. Processing times of 0.1-0.17 seconds per 640×640 pixel image enable real-time deployment. The system demonstrated a mean absolute error of 6.81 eggs with 46.8% perfect predictions on validation data. This research validates synthetic data effectiveness for tiny biological object detection while providing practical automation tools for global vector surveillance programs, potentially transforming mosquito monitoring capabilities essential for disease outbreak prevention. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Creative Technology BSc (50447) |
| Link to this item: | https://purl.utwente.nl/essays/108056 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page