Detection of the mite Varroa destructor in honey bee cells by video sequence processing

Abstract

It is presented an algorithm for detecting and tracking a mite, from videos provided by the Centro de Investigación Apícola Tropical (CINAT-Costa Rica). These registrations correspond to the presence of the Varroa destructor mite in an Africanized honey bee cell, in a controlled environment. The main objective in this paper is to present the various stages of development of the algorithm and show the results obtained in the percentage of success in detecting. Therefore, we have implemented a calibration system in order to have a frame more enhanced versus the original video. This calibration is done by the searching of the Movement Active Area and definition of our object (the mite V destructor), and finally, an automatic detection and tracking are done. We have reached up to 93.75% for right detection and tracking, working on real time under a Matlab environment.