Research library

An AI-Based Open-Source Software for Varroa Mite Fall Analysis in Honeybee Colonies

Infestation by *Varroa destructor* is responsible for high mortality rates in *Apis mellifera* colonies worldwide. This study developed and field-tested a new free software (VarroDetector) based on a deep learning approach for the automated detection and counting of *Varroa* mites using smartphone images of sticky boards collected in honeybee colonies. A total of 204 sheets were collected, divided into four frames using green strings, and photographed under controlled lighting conditions with different smartphone models at a minimum resolution of 48 megapixels. The *Varroa* detection algorithm comprises two main steps: first, the region of interest where *Varroa* mites must be counted is established; then a one-stage detector (YOLO v11 Nano) is applied. The results obtained with VarroDetector were highly correlated with manual control counts (R² = 0.98–0.99, depending on the smartphone camera used). When *Varroa* mite numbers were higher than 50 per sheet, VarroDetector outperformed trained human visual inspectors in reliability while significantly reducing processing time. The software runs offline on a smartphone and is released as open source.