Detailled semantic labeling (2D) result

Overall statistics, reference set: full_reference Number of non processed tiles by this participant: 0 ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.933 0.013 0.035 0.012 0.001 0.007 building 0.011 0.977 0.005 0.002 0.002 0.003 low_veg 0.044 0.012 0.900 0.038 0.000 0.006 tree 0.042 0.008 0.159 0.788 0.001 0.002 car 0.144 0.002 0.003 0.018 0.818 0.014 clutter 0.308 0.113 0.187 0.019 0.009 0.365 Precision/Correctness 0.891 0.947 0.798 0.909 0.931 0.779 Recall/Completeness 0.933 0.977 0.900 0.788 0.818 0.365 F1 0.911 0.962 0.846 0.844 0.871 0.497 Overall accuracy 0.883	Overall statistics, reference set: no_boundary Number of non processed tiles by this participant: 0 ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.948 0.009 0.028 0.008 0.000 0.006 building 0.009 0.983 0.003 0.001 0.002 0.002 low_veg 0.034 0.008 0.922 0.031 0.000 0.006 tree 0.031 0.006 0.142 0.817 0.001 0.002 car 0.023 0.002 0.001 0.013 0.945 0.016 clutter 0.308 0.096 0.174 0.016 0.009 0.397 Precision/Correctness 0.914 0.963 0.821 0.932 0.938 0.795 Recall/Completeness 0.948 0.983 0.922 0.817 0.945 0.397 F1 0.931 0.973 0.868 0.871 0.941 0.530 Overall accuracy 0.906
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8956 0.0195 0.0555 0.0081 0.0007 0.0206 building 0.0061 0.9801 0.0063 0.0037 0.0001 0.0037 low_veg 0.0230 0.0364 0.9058 0.0254 0.0002 0.0092 tree 0.0373 0.0150 0.1667 0.7771 0.0008 0.0030 car 0.1299 0.0058 0.0031 0.0107 0.7951 0.0554 clutter 0.1137 0.2341 0.1708 0.0191 0.0020 0.4602 Precision/Correctness 0.915 0.883 0.828 0.943 0.952 0.473 Recall/Completeness 0.896 0.980 0.906 0.777 0.795 0.460 F1 0.905 0.929 0.865 0.852 0.866 0.466 Overall accuracy 0.877 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_2_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9153 0.0136 0.0469 0.0045 0.0001 0.0197 building 0.0038 0.9872 0.0028 0.0030 0.0001 0.0031 low_veg 0.0165 0.0348 0.9229 0.0173 0.0001 0.0084 tree 0.0284 0.0130 0.1448 0.8108 0.0005 0.0024 car 0.0353 0.0058 0.0016 0.0053 0.8839 0.0681 clutter 0.0735 0.2359 0.1282 0.0126 0.0022 0.5476 Precision/Correctness 0.942 0.903 0.854 0.964 0.974 0.460 Recall/Completeness 0.915 0.987 0.923 0.811 0.884 0.548 F1 0.929 0.943 0.887 0.881 0.927 0.500 Overall accuracy 0.901 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_2_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9169 0.0075 0.0675 0.0068 0.0003 0.0010 building 0.0022 0.9883 0.0065 0.0020 0.0000 0.0009 low_veg 0.0226 0.0032 0.9536 0.0186 0.0000 0.0021 tree 0.0227 0.0053 0.2200 0.7502 0.0006 0.0012 car 0.1195 0.0035 0.0023 0.0079 0.8601 0.0067 clutter 0.0869 0.3570 0.1881 0.0415 0.0069 0.3196 Precision/Correctness 0.884 0.923 0.815 0.967 0.947 0.426 Recall/Completeness 0.917 0.988 0.954 0.750 0.860 0.320 F1 0.900 0.955 0.879 0.845 0.901 0.365 Overall accuracy 0.875 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_2_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9414 0.0040 0.0504 0.0035 0.0000 0.0007 building 0.0006 0.9946 0.0029 0.0011 0.0000 0.0008 low_veg 0.0184 0.0012 0.9649 0.0137 0.0000 0.0018 tree 0.0181 0.0043 0.2020 0.7740 0.0004 0.0011 car 0.0133 0.0022 0.0003 0.0042 0.9717 0.0083 clutter 0.0436 0.4104 0.1584 0.0458 0.0071 0.3346 Precision/Correctness 0.907 0.949 0.832 0.978 0.964 0.393 Recall/Completeness 0.941 0.995 0.965 0.774 0.972 0.335 F1 0.924 0.971 0.893 0.864 0.968 0.361 Overall accuracy 0.893 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_3_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9234 0.0205 0.0396 0.0119 0.0007 0.0039 building 0.0058 0.9870 0.0031 0.0025 0.0000 0.0015 low_veg 0.0572 0.0191 0.8875 0.0333 0.0000 0.0029 tree 0.0372 0.0104 0.1374 0.8125 0.0014 0.0012 car 0.1315 0.0036 0.0015 0.0197 0.8324 0.0113 clutter 0.3064 0.1794 0.2590 0.0136 0.0043 0.2373 Precision/Correctness 0.876 0.937 0.808 0.940 0.959 0.694 Recall/Completeness 0.923 0.987 0.887 0.812 0.832 0.237 F1 0.899 0.962 0.846 0.872 0.891 0.354 Overall accuracy 0.887 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_3_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9469 0.0133 0.0294 0.0070 0.0000 0.0033 building 0.0041 0.9918 0.0009 0.0018 0.0000 0.0014 low_veg 0.0467 0.0092 0.9172 0.0248 0.0000 0.0021 tree 0.0279 0.0085 0.1185 0.8433 0.0009 0.0008 car 0.0209 0.0039 0.0007 0.0136 0.9482 0.0128 clutter 0.3207 0.1782 0.2264 0.0072 0.0052 0.2623 Precision/Correctness 0.907 0.960 0.839 0.961 0.977 0.703 Recall/Completeness 0.947 0.992 0.917 0.843 0.948 0.262 F1 0.926 0.976 0.876 0.898 0.962 0.382 Overall accuracy 0.915 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_3_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9007 0.0181 0.0601 0.0147 0.0009 0.0054 building 0.0065 0.9852 0.0040 0.0017 0.0000 0.0025 low_veg 0.0283 0.0104 0.9298 0.0291 0.0000 0.0023 tree 0.0279 0.0054 0.2078 0.7558 0.0009 0.0021 car 0.1263 0.0070 0.0013 0.0308 0.8239 0.0107 clutter 0.2255 0.1942 0.1844 0.0287 0.0043 0.3628 Precision/Correctness 0.885 0.939 0.776 0.939 0.945 0.785 Recall/Completeness 0.901 0.985 0.930 0.756 0.824 0.363 F1 0.893 0.962 0.846 0.838 0.880 0.496 Overall accuracy 0.873 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_3_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9275 0.0112 0.0463 0.0097 0.0002 0.0051 building 0.0054 0.9898 0.0015 0.0011 0.0000 0.0022 low_veg 0.0192 0.0047 0.9530 0.0214 0.0000 0.0017 tree 0.0218 0.0040 0.1892 0.7826 0.0007 0.0016 car 0.0198 0.0068 0.0004 0.0216 0.9394 0.0120 clutter 0.2153 0.1884 0.1487 0.0242 0.0049 0.4184 Precision/Correctness 0.915 0.960 0.800 0.959 0.960 0.807 Recall/Completeness 0.927 0.990 0.953 0.783 0.939 0.418 F1 0.921 0.975 0.870 0.862 0.950 0.551 Overall accuracy 0.898 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_4_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9441 0.0076 0.0321 0.0146 0.0002 0.0014 building 0.0090 0.9800 0.0053 0.0040 0.0000 0.0016 low_veg 0.0397 0.0062 0.8841 0.0689 0.0000 0.0011 tree 0.0370 0.0028 0.0662 0.8928 0.0009 0.0004 car 0.1585 0.0007 0.0038 0.0272 0.8027 0.0070 clutter 0.5282 0.1172 0.1831 0.0510 0.0043 0.1161 Precision/Correctness 0.844 0.966 0.798 0.852 0.973 0.856 Recall/Completeness 0.944 0.980 0.884 0.893 0.803 0.116 F1 0.891 0.973 0.839 0.872 0.880 0.204 Overall accuracy 0.882 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_4_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9599 0.0043 0.0252 0.0094 0.0001 0.0011 building 0.0055 0.9881 0.0020 0.0029 0.0000 0.0015 low_veg 0.0297 0.0017 0.9088 0.0590 0.0000 0.0008 tree 0.0234 0.0014 0.0516 0.9228 0.0005 0.0003 car 0.0180 0.0002 0.0003 0.0197 0.9539 0.0079 clutter 0.5567 0.1033 0.1646 0.0430 0.0049 0.1275 Precision/Correctness 0.871 0.977 0.835 0.886 0.976 0.870 Recall/Completeness 0.960 0.988 0.909 0.923 0.954 0.128 F1 0.913 0.983 0.870 0.904 0.965 0.222 Overall accuracy 0.907 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_4_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8904 0.0218 0.0676 0.0097 0.0003 0.0102 building 0.0048 0.9855 0.0035 0.0033 0.0000 0.0028 low_veg 0.0753 0.0158 0.8435 0.0415 0.0000 0.0239 tree 0.0396 0.0062 0.1625 0.7835 0.0014 0.0068 car 0.1502 0.0045 0.0030 0.0217 0.7996 0.0209 clutter 0.4825 0.0463 0.1485 0.0074 0.0004 0.3149 Precision/Correctness 0.747 0.930 0.754 0.914 0.979 0.750 Recall/Completeness 0.890 0.986 0.844 0.784 0.800 0.315 F1 0.812 0.957 0.796 0.844 0.880 0.443 Overall accuracy 0.819 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_4_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9049 0.0164 0.0627 0.0062 0.0000 0.0098 building 0.0029 0.9906 0.0016 0.0024 0.0000 0.0024 low_veg 0.0635 0.0099 0.8670 0.0347 0.0000 0.0249 tree 0.0304 0.0045 0.1423 0.8153 0.0011 0.0065 car 0.0309 0.0041 0.0013 0.0159 0.9246 0.0232 clutter 0.4963 0.0419 0.1313 0.0053 0.0002 0.3251 Precision/Correctness 0.762 0.948 0.775 0.935 0.987 0.762 Recall/Completeness 0.905 0.991 0.867 0.815 0.925 0.325 F1 0.827 0.969 0.819 0.871 0.955 0.456 Overall accuracy 0.839 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_4_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9402 0.0138 0.0393 0.0030 0.0008 0.0029 building 0.0027 0.9887 0.0073 0.0008 0.0000 0.0006 low_veg 0.0550 0.0213 0.9042 0.0170 0.0001 0.0025 tree 0.0882 0.0119 0.1735 0.7185 0.0046 0.0033 car 0.1612 0.0027 0.0029 0.0075 0.8157 0.0099 clutter 0.3659 0.2403 0.1407 0.0107 0.0033 0.2391 Precision/Correctness 0.900 0.948 0.779 0.953 0.952 0.756 Recall/Completeness 0.940 0.989 0.904 0.719 0.816 0.239 F1 0.920 0.968 0.837 0.819 0.879 0.363 Overall accuracy 0.897 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_4_15_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9548 0.0071 0.0338 0.0017 0.0003 0.0023 building 0.0010 0.9926 0.0056 0.0004 0.0000 0.0004 low_veg 0.0363 0.0122 0.9372 0.0124 0.0000 0.0019 tree 0.0699 0.0097 0.1521 0.7623 0.0033 0.0028 car 0.0275 0.0021 0.0003 0.0038 0.9538 0.0125 clutter 0.3829 0.2425 0.1205 0.0082 0.0028 0.2431 Precision/Correctness 0.930 0.966 0.804 0.971 0.969 0.763 Recall/Completeness 0.955 0.993 0.937 0.762 0.954 0.243 F1 0.942 0.979 0.866 0.854 0.961 0.369 Overall accuracy 0.923 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_5_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9345 0.0140 0.0335 0.0146 0.0002 0.0030 building 0.0063 0.9872 0.0035 0.0021 0.0000 0.0009 low_veg 0.0697 0.0111 0.8662 0.0512 0.0000 0.0019 tree 0.0570 0.0059 0.1484 0.7862 0.0017 0.0008 car 0.1701 0.0008 0.0008 0.0105 0.8020 0.0159 clutter 0.2306 0.2641 0.1919 0.0142 0.0539 0.2453 Precision/Correctness 0.902 0.946 0.819 0.859 0.910 0.788 Recall/Completeness 0.935 0.987 0.866 0.786 0.802 0.245 F1 0.918 0.966 0.842 0.821 0.853 0.374 Overall accuracy 0.892 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_5_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9505 0.0083 0.0287 0.0102 0.0000 0.0023 building 0.0032 0.9926 0.0019 0.0015 0.0000 0.0008 low_veg 0.0627 0.0062 0.8860 0.0438 0.0000 0.0013 tree 0.0401 0.0038 0.1349 0.8193 0.0013 0.0006 car 0.0317 0.0006 0.0001 0.0083 0.9409 0.0185 clutter 0.1977 0.2518 0.1852 0.0125 0.0697 0.2832 Precision/Correctness 0.931 0.965 0.845 0.887 0.913 0.804 Recall/Completeness 0.951 0.993 0.886 0.819 0.941 0.283 F1 0.941 0.978 0.865 0.852 0.927 0.419 Overall accuracy 0.917 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_5_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9452 0.0049 0.0230 0.0194 0.0002 0.0073 building 0.0088 0.9858 0.0043 0.0008 0.0001 0.0002 low_veg 0.0799 0.0049 0.8489 0.0563 0.0000 0.0100 tree 0.0520 0.0065 0.1090 0.8267 0.0019 0.0038 car 0.1768 0.0004 0.0102 0.0204 0.7702 0.0219 clutter 0.2264 0.1908 0.1901 0.0400 0.0022 0.3504 Precision/Correctness 0.841 0.933 0.812 0.876 0.963 0.834 Recall/Completeness 0.945 0.986 0.849 0.827 0.770 0.350 F1 0.890 0.959 0.830 0.851 0.856 0.493 Overall accuracy 0.867 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_5_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9610 0.0021 0.0159 0.0143 0.0001 0.0065 building 0.0068 0.9903 0.0022 0.0005 0.0000 0.0001 low_veg 0.0668 0.0019 0.8705 0.0512 0.0000 0.0096 tree 0.0406 0.0053 0.0983 0.8513 0.0013 0.0032 car 0.0305 0.0003 0.0095 0.0143 0.9216 0.0238 clutter 0.2128 0.1628 0.1867 0.0396 0.0022 0.3959 Precision/Correctness 0.872 0.953 0.835 0.898 0.972 0.852 Recall/Completeness 0.961 0.990 0.871 0.851 0.922 0.396 F1 0.914 0.971 0.852 0.874 0.946 0.541 Overall accuracy 0.892 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_5_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9325 0.0069 0.0448 0.0131 0.0005 0.0022 building 0.0087 0.9792 0.0049 0.0034 0.0001 0.0037 low_veg 0.0414 0.0055 0.8668 0.0752 0.0001 0.0111 tree 0.0404 0.0050 0.1049 0.8472 0.0013 0.0012 car 0.1793 0.0008 0.0032 0.0169 0.7851 0.0147 clutter 0.2977 0.1176 0.1246 0.0138 0.0206 0.4256 Precision/Correctness 0.907 0.951 0.756 0.844 0.919 0.885 Recall/Completeness 0.933 0.979 0.867 0.847 0.785 0.426 F1 0.919 0.965 0.807 0.845 0.847 0.575 Overall accuracy 0.885 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_5_15_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9454 0.0034 0.0396 0.0094 0.0004 0.0017 building 0.0055 0.9858 0.0029 0.0024 0.0001 0.0034 low_veg 0.0270 0.0018 0.8912 0.0694 0.0000 0.0105 tree 0.0287 0.0029 0.0921 0.8742 0.0010 0.0011 car 0.0307 0.0006 0.0002 0.0142 0.9371 0.0172 clutter 0.2972 0.0822 0.1137 0.0117 0.0230 0.4723 Precision/Correctness 0.930 0.972 0.779 0.871 0.918 0.898 Recall/Completeness 0.945 0.986 0.891 0.874 0.937 0.472 F1 0.938 0.979 0.831 0.873 0.927 0.619 Overall accuracy 0.909 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_6_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9561 0.0058 0.0193 0.0133 0.0002 0.0053 building 0.0307 0.9410 0.0111 0.0010 0.0103 0.0060 low_veg 0.0164 0.0028 0.9605 0.0196 0.0000 0.0006 tree 0.0707 0.0029 0.1156 0.8090 0.0009 0.0009 car 0.1364 0.0002 0.0012 0.0059 0.8478 0.0083 clutter 0.3365 0.0441 0.3707 0.0214 0.0046 0.2228 Precision/Correctness 0.927 0.982 0.879 0.903 0.791 0.546 Recall/Completeness 0.956 0.941 0.961 0.809 0.848 0.223 F1 0.941 0.961 0.918 0.853 0.818 0.316 Overall accuracy 0.920 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_6_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9690 0.0035 0.0149 0.0083 0.0000 0.0043 building 0.0271 0.9486 0.0076 0.0006 0.0105 0.0056 low_veg 0.0095 0.0012 0.9736 0.0152 0.0000 0.0005 tree 0.0517 0.0019 0.1026 0.8425 0.0006 0.0007 car 0.0174 0.0002 0.0008 0.0040 0.9676 0.0101 clutter 0.3344 0.0435 0.3793 0.0204 0.0041 0.2183 Precision/Correctness 0.945 0.989 0.897 0.931 0.778 0.546 Recall/Completeness 0.969 0.949 0.974 0.842 0.968 0.218 F1 0.957 0.968 0.934 0.884 0.863 0.312 Overall accuracy 0.937 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_6_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9681 0.0045 0.0172 0.0076 0.0009 0.0017 building 0.0346 0.9497 0.0040 0.0011 0.0090 0.0016 low_veg 0.0534 0.0060 0.8957 0.0425 0.0001 0.0022 tree 0.0466 0.0078 0.1094 0.8335 0.0020 0.0007 car 0.1099 0.0003 0.0016 0.0164 0.8667 0.0052 clutter 0.2641 0.1081 0.2176 0.0155 0.0163 0.3785 Precision/Correctness 0.921 0.980 0.855 0.881 0.882 0.836 Recall/Completeness 0.968 0.950 0.896 0.833 0.867 0.379 F1 0.944 0.964 0.875 0.856 0.874 0.521 Overall accuracy 0.922 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_6_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9787 0.0017 0.0128 0.0053 0.0001 0.0014 building 0.0306 0.9564 0.0019 0.0006 0.0091 0.0014 low_veg 0.0386 0.0022 0.9212 0.0364 0.0000 0.0016 tree 0.0337 0.0066 0.0944 0.8635 0.0012 0.0006 car 0.0073 0.0001 0.0002 0.0119 0.9745 0.0059 clutter 0.2416 0.1165 0.2028 0.0116 0.0167 0.4108 Precision/Correctness 0.942 0.987 0.880 0.908 0.878 0.856 Recall/Completeness 0.979 0.956 0.921 0.864 0.974 0.411 F1 0.960 0.971 0.900 0.885 0.924 0.555 Overall accuracy 0.940 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_6_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9358 0.0039 0.0326 0.0056 0.0011 0.0211 building 0.0142 0.9729 0.0065 0.0009 0.0000 0.0054 low_veg 0.0395 0.0025 0.9177 0.0210 0.0003 0.0190 tree 0.0516 0.0132 0.1951 0.7358 0.0019 0.0023 car 0.1228 0.0011 0.0058 0.0206 0.8397 0.0100 clutter 0.1692 0.0115 0.2180 0.0084 0.0102 0.5828 Precision/Correctness 0.919 0.980 0.704 0.946 0.949 0.733 Recall/Completeness 0.936 0.973 0.918 0.736 0.840 0.583 F1 0.927 0.976 0.797 0.828 0.891 0.649 Overall accuracy 0.887 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_6_15_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9494 0.0015 0.0254 0.0031 0.0000 0.0206 building 0.0110 0.9800 0.0039 0.0005 0.0000 0.0046 low_veg 0.0249 0.0007 0.9410 0.0148 0.0000 0.0186 tree 0.0409 0.0120 0.1766 0.7673 0.0013 0.0018 car 0.0204 0.0010 0.0011 0.0162 0.9497 0.0116 clutter 0.1631 0.0100 0.2070 0.0069 0.0105 0.6024 Precision/Correctness 0.941 0.987 0.723 0.966 0.966 0.748 Recall/Completeness 0.949 0.980 0.941 0.767 0.950 0.602 F1 0.945 0.983 0.818 0.855 0.958 0.667 Overall accuracy 0.907 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_7_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9270 0.0300 0.0117 0.0228 0.0004 0.0081 building 0.0069 0.9814 0.0022 0.0015 0.0000 0.0080 low_veg 0.0536 0.0178 0.8189 0.1084 0.0000 0.0012 tree 0.0255 0.0061 0.1459 0.8187 0.0002 0.0037 car 0.1087 0.0088 0.0008 0.0330 0.8388 0.0099 clutter 0.2092 0.0201 0.2050 0.0017 0.0053 0.5586 Precision/Correctness 0.940 0.920 0.696 0.733 0.945 0.888 Recall/Completeness 0.927 0.981 0.819 0.819 0.839 0.559 F1 0.934 0.950 0.752 0.773 0.889 0.686 Overall accuracy 0.886 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_7_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9401 0.0264 0.0094 0.0162 0.0001 0.0078 building 0.0048 0.9856 0.0013 0.0008 0.0000 0.0074 low_veg 0.0424 0.0144 0.8421 0.1002 0.0000 0.0009 tree 0.0165 0.0041 0.1393 0.8361 0.0000 0.0040 car 0.0097 0.0109 0.0001 0.0256 0.9418 0.0120 clutter 0.1999 0.0185 0.2056 0.0014 0.0053 0.5693 Precision/Correctness 0.950 0.931 0.704 0.770 0.950 0.895 Recall/Completeness 0.940 0.986 0.842 0.836 0.942 0.569 F1 0.945 0.958 0.767 0.801 0.946 0.696 Overall accuracy 0.899 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image