=============================================================
↓ predicted || reference →
 power
low_veg
imp_surf
car
fence_hedge
roof
fac
shrub
tree
missing points
power
0.743
0.000
0.000
0.000
0.008
0.000
0.038
0.000
0.210
0
low_veg
0.138
0.045
0.127
0.021
0.094
0.310
0.055
0.052
0.158
0
imp_surf
0.041
0.034
0.543
0.018
0.015
0.292
0.026
0.004
0.027
0
car
0.467
0.008
0.013
0.221
0.125
0.106
0.004
0.012
0.045
0
fence_hedge
0.271
0.006
0.006
0.067
0.217
0.046
0.022
0.140
0.227
0
roof
0.142
0.005
0.026
0.003
0.023
0.561
0.103
0.007
0.131
0
fac
0.095
0.000
0.000
0.000
0.006
0.044
0.505
0.006
0.343
0
shrub
0.206
0.007
0.002
0.026
0.166
0.089
0.021
0.216
0.268
0
tree
0.135
0.001
0.000
0.002
0.024
0.048
0.068
0.058
0.666
0
Precision/Correctness
 0.009
0.506
0.782
0.131
0.078
0.479
0.193
0.336
0.444
Recall/Completeness
 0.743
0.045
0.543
0.221
0.217
0.561
0.505
0.216
0.666
F1
 0.017
0.082
0.641
0.165
0.114
0.517
0.280
0.263
0.533

Overall accuracy 0.415

Data Download


Download a zip file containing laz files per class. Those can be visualized with CloudCompare

Click here for download!

Data previews (derived from voxelisation at 50cm voxelcube)

Participant labeling (lowest points when multiple labels at one X/Y position)

Red/green image, indicating wrongly classified areas

!!! All red/green visualisations (also the point clouds in the zip file) are defined from the reference point of view.