Table 1 Experimental data preparation.
From: Individual honey bee tracking in a beehive environment using deep learning and Kalman filter
Data Characteristic | Mask R-CNN (Training) | Mask R-CNN + Kalman (Testing) |
|---|---|---|
Image frame size | 1920 × 1080 pixels | 1920 × 1080 pixels |
Number of images | Training Set: 35 frames | – |
Validate Set: 5 frames | ||
Test Set: | ||
rames (at frame rate 5 fps) | ||
100 frames (at frame rate 10 fps) | ||
150 frames (at frame rate 15 fps) | ||
Frame rate (fps) | – | 5 fps, 10 fps, 15 fps |
Sequence length | – | 10 s/sequence |
Number of image sequences | – | 50 frames (at frame rate 5 fps) |
100 frames (at frame rate 10 fps) | ||
150 frames (at frame rate 15 fps) | ||
Number of bees per image frame | ≈ 370 bees | ≈ 370 bees |
