| |
Abstract |
Varroa destructor is a parasite harming bee colonies. As the worldwide bee population is in danger, beekeepers as well as researchers are looking for methods to monitor the health of bee hives. In this context, we present a preliminary study to detect parasites on bee videos by means of image analysis and machine learning techniques. For this purpose, each video frame is analyzed individually to extract bee image patches, which are then processed to compute image descriptors and finally classified into mite and no mite bees. The experimental results demonstrated the adequacy of the proposed method, which will be a perfect stepping stone for a further bee monitoring system. |
|
