OpenDroneMap / ODMSemantic3D

An open photogrammetry dataset of classified 3D point clouds for automated semantic segmentation. CC BY-SA 4.0
Other
15 stars 7 forks source link

Added toledo point cloud #18

Closed HeDo88TH closed 1 year ago

u4gbot commented 1 year ago
Comparison for brighton_beach_small (Differences) Accuracy: 94.11% vs 96.30% | Label | Accuracy | Recall | Precision | F1 | | -------------------- | ---------- | ---------- | ---------- | ---------- | | | 94.06% | 94.12% | 99.93% | 0.97 | | ground | 96.50% | 98.51% | 97.93% | 0.98 | | | 2.44% | 4.39% | -2.00% | 0.01 | | | | | | | | | 66.95% | 95.07% | 69.36% | 0.80 | | low_vegetation | 37.76% | 49.69% | 61.14% | 0.55 | | | -29.19% | -45.38% | -8.22% | -0.25 | | | | | | | | | 45.35% | 83.98% | 49.64% | 0.62 | | human_made_object | 50.90% | 51.49% | 97.82% | 0.67 | | | 5.56% | -32.50% | 48.18% | 0.05 | | | | | | | | | 14.64% | 99.56% | 14.65% | 0.26 | | building | 21.62% | 44.65% | 29.53% | 0.36 | | | 6.98% | -54.91% | 14.88% | 0.10 | | | | | | |
Comparison for sheffield_park_small (Differences) Accuracy: 97.85% vs 26.67% | Label | Accuracy | Recall | Precision | F1 | | -------------------- | ---------- | ---------- | ---------- | ---------- | | | 98.04% | 98.33% | 99.70% | 0.99 | | ground | 9.40% | 9.72% | 73.91% | 0.17 | | | -88.64% | -88.61% | -25.79% | -0.82 | | | | | | | | | 95.84% | 98.10% | 97.66% | 0.98 | | low_vegetation | 19.42% | 99.97% | 19.42% | 0.33 | | | -76.43% | 1.88% | -78.24% | -0.65 | | | | | | | | | 34.76% | 94.35% | 35.50% | 0.52 | | human_made_object | 3.75% | 4.97% | 13.27% | 0.07 | | | -31.01% | -89.38% | -22.23% | -0.44 | | | | | | | | | 92.46% | 94.00% | 98.27% | 0.96 | | building | 26.33% | 30.57% | 65.51% | 0.42 | | | -66.13% | -63.42% | -32.76% | -0.54 | | | | | | |

Average (2 datasets)

Accuracy: 95.98% vs 61.48%

Label Accuracy Recall Precision F1
96.05% 96.22% 99.82% 0.98
ground 52.95% 54.11% 85.92% 0.58
-43.10% -42.11% -13.90% -0.40
81.40% 96.58% 83.51% 0.89
low_vegetation 28.59% 74.83% 40.28% 0.44
-52.81% -21.75% -43.23% -0.45
40.06% 89.17% 42.57% 0.57
human_made_object 27.33% 28.23% 55.55% 0.37
-12.73% -60.94% 12.98% -0.20
53.55% 96.78% 56.46% 0.61
building 23.97% 37.61% 47.52% 0.39
-29.58% -59.17% -8.94% -0.22