Varroa Mite Detection in Honey Bees with Artificial Vision

The preservation of species is beneficial for the subsistence of life on planet earth. The honey bee, considered a pollinating and food-producing species for humans, faces the attack of the Varroa parasite; Several solutions to detect, exterminate and contain this ectoparasite in bee colonies have been proposed. This paper presents and describes the development of a non-invasive system with artificial vision, which uses Distributed Computing and the Internet of Things to recognize and analyze the thorax and abdomen of honey bees, in images acquired in the entrances of the hives, for the early detection of the Varroa mite. Two techniques have been experimented with for the recognition of the bee at the entrance of the hive, the Support Vector Machine (SVM) and k-Nearest Neighbors (kNN), and with the same two techniques to analyze the Varroa mite in the thorax and neck. bee abdomen. The experiments carried out and the results obtained with each technique are described.