Selected RGB Frame
The tray frame used as the visual base for the fused Arabidopsis analysis.
Tray showcase
This popup shows a full tray analysis where neighboring rosettes are touching, making instance separation, convex hull fitting, and trait attribution substantially harder than in clean benchmark trays.
The tray frame used as the visual base for the fused Arabidopsis analysis.
Plant-level instance separation projected back onto the tray despite neighboring rosettes touching.
Ellipse fits, convex hulls, and bounding geometry used to keep per-plant shapes stable under overlap.
Integer plant labels for the full tray, useful for checking instance consistency across a crowded canopy.
Rosette size spread across the tray, highlighting how overlap shifts apparent canopy size.
ARKit-referenced height estimates relative to the tray baseline for the full Arabidopsis tray.
Bubble chart showing how canopy area and 3D height interact across the touching rosette tray.
Full tray export from the package, including leaf count estimates, area, height, canopy hull area, voxel volume, solidity, circularity, and LiDAR support for each detected plant.
| Plant | Leaves | Area (px²) | Height (m) | Canopy area (m²) | Voxel volume (m³) | Solidity | Circularity | LiDAR points |
|---|---|---|---|---|---|---|---|---|
| 1 | 8 | 32597 | 0.0 | 0.0 | 0.0 | 0.6375 | 0.1238 | 0 |
| 2 | 6 | 25299 | 0.0 | 0.0 | 0.0 | 0.4822 | 0.0942 | 0 |
| 3 | 7 | 33347 | 0.1192 | 0.001018 | 7.4e-06 | 0.4605 | 0.0944 | 59 |
| 4 | 7 | 32488 | 0.1205 | 0.001492 | 2.73e-05 | 0.6312 | 0.1603 | 222 |
| 5 | 9 | 35159 | 0.1182 | 0.002555 | 2.44e-05 | 0.5581 | 0.0737 | 261 |
| 6 | 8 | 38267 | 0.1193 | 0.004708 | 4.18e-05 | 0.4701 | 0.0867 | 341 |
| 7 | 7 | 28115 | 0.119 | 0.004025 | 3.54e-05 | 0.4156 | 0.092 | 287 |
| 8 | 11 | 38164 | 0.1187 | 0.003987 | 5.09e-05 | 0.5613 | 0.0804 | 413 |
| 9 | 10 | 41810 | 0.1182 | 0.005967 | 3.81e-05 | 0.5809 | 0.0877 | 399 |
| 10 | 9 | 42502 | 0.1182 | 0.004835 | 5.06e-05 | 0.4905 | 0.0713 | 413 |
| 11 | 11 | 32242 | 0.1181 | 0.004824 | 3.03e-05 | 0.333 | 0.0637 | 246 |
| 12 | 6 | 30816 | 0.1196 | 0.004815 | 4.23e-05 | 0.4881 | 0.111 | 342 |
| 13 | 9 | 38288 | 0.1198 | 0.004933 | 3.06e-05 | 0.5072 | 0.103 | 335 |
| 14 | 10 | 42865 | 0.1186 | 0.00383 | 4.15e-05 | 0.5134 | 0.0754 | 340 |
| 15 | 10 | 35460 | 0.119 | 0.003351 | 3.05e-05 | 0.4727 | 0.0669 | 252 |
| 16 | 9 | 35480 | 0.1219 | 0.008571 | 5.96e-05 | 0.4799 | 0.1026 | 489 |
| 17 | 8 | 41763 | 0.1226 | 0.007072 | 8.18e-05 | 0.6266 | 0.0843 | 773 |
| 18 | 9 | 44337 | 0.1217 | 0.005823 | 9.01e-05 | 0.6624 | 0.1247 | 734 |
| 19 | 8 | 34746 | 0.1187 | 0.005093 | 8.36e-05 | 0.5879 | 0.0965 | 673 |
| 20 | 11 | 43850 | 0.1226 | 0.008864 | 0.0001456 | 0.6352 | 0.1118 | 1192 |
Heights use the ARKit world Y-axis relative to the tray baseline, voxel volume is an occupancy-based proxy, solidity is area divided by convex hull area, and leaf counts are peak-based estimates derived from rosette-mask topology.