Status: about to be finished
Supervisor: Martin Kampel
Start: as soon as possible
In the project MIC-Cam a system for the automated detection of mites was developed. It monitors the bees by camera when departing and landing at the flight board of the bee hive to recognise attached mites in real time without restricting the bees. We apply innovative methods of digital image processing and machine learning implemented on single board processors for a robust and real-time detection of mites in the image sequences.
No we would like to extend the functionality by recognizing polls of the flying in bee, and furthermore we also would like to measure the quantity of polls carried by the incoming bees.
The goal is to develop a solution for bee poll detection. One straight forward approach would be to investigate color features of the poll in order to detect them. A deep Learning strategy might by appropriate as well (see  for an example). Existing video data has to be investigated in order to find sample images for training or segmentation.
If you are interested, send me an email: Martin Kampel
 I. F. Rodriguez, R. Megret, E. Acuna, J. L. Agosto-Rivera and T. Giray, “Recognition of Pollen-Bearing Bees from Video Using Convolutional Neural Network,” 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), Lake Tahoe, NV, 2018, pp. 314-322.