From: Detecting and tracking honeybees in 3D at the beehive entrance using stereo vision
Reference | Targets | Loc/obs | Cam/fps | Detection/likehood method | Tracking method |
---|---|---|---|---|---|
[8] | < 15 ants | 2D/2D | 1/30 | ABS | Basic GNN |
[19] | < 20 ants | 2D/2D | 1/30 | Appearance model | MRF-augmented FP |
[15] | 1 bee* | 2D/2D | 1/15 | Eigenbee (PCA) | PF |
[16] | 1 bee* | 2D/2D | 1/15 | Adaptive appearance model (e.g., color image) | PF supported by a behavior model |
[17] | 1 bee* | 2D/2D | 1/15 | Weighted adaptive appearance model (e.g., color image) | Idem + geometric constraints |
[13] | < 100 bees | 2D/2D | 1/30 | Vector quantization (VQ) | Overlapping ellipse |
[18] | < 100 bees | 2D/2D | 1/14 | Tag detection (method not mentioned) | Mean shift |
[11] | 1 bee | 2D/2D | 1/? | Viola-Jones detector | Combined NN classification of BOF |
[9] | n bees | 3D/2D | 1/30 | ABS + ellipse matching | Basic GNN |
[10] | 1 bee | 3D/2D | 1/24 | ABS | NN |
[14] | < 100 bats | 3D/3D | 3 IR/125 | Direct linear transform | MHT |
Our | < 25 bees | 3D/3D | 2/60 | Hybrid intensity/depth segmentation | GNN and 3D (re)projection |