期刊
COMPUTERS AND ELECTRONICS IN AGRICULTURE
卷 163, 期 -, 页码 -出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.compag.2019.05.050
关键词
Object detection; Multiple object tracking; Image processing; Embedded system; Imidacloprid pesticide
资金
- Ministry of Science and Technology, ROC [104-2627-M-002014, 105-2627-M-002-012, 106-2627-M-002-004]
This study presents a real-time imaging system for monitoring honey bee activity by counting the honey bees entering and exiting the beehive. Images are continuously acquired at the beehive entrance, and the honey bees in the images are segmented and detected using the background subtraction method. Tracking of honey bees is achieved using an integrated Kalman Filter and Hungarian algorithm tracking method. The tracking algorithm was used to determine the incoming and outgoing activity of individual honey bees with an automatic counting accuracy of 93.9 +/- 1.1% in comparison with manual counting. The in-and-out activity and environmental information from sensors were transmitted to a remote server via a 4G LTE router for later analyses. To evaluate and demonstrate the performance of the system, the imaging systems were installed in multiple beehives, and designed experiments were carried out. The experiments involved long-term monitoring to assess the pesticide effects on honey bee behaviors by treating the beehives with different concentrations of imidacloprid pesticides. Using the in-and-out activities information collected, indices were further derived and analyzed. The information obtained from the experiments was used to assess the health conditions of the honey bee colonies monitored. Different measurement results such as hourly and daily frequencies of honey bees' in-and-out activity under different conditions were presented. The developed imaging system is cost effective and can be used as a useful tool for long-term monitoring of the health and activity of honey bee colonies and to further analyze behavior traits of honey bees.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据