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 0.899 0.035 0.036 0.018 0.003 0.009 building 0.018 0.961 0.012 0.004 0.000 0.004 low_veg 0.047 0.022 0.860 0.062 0.000 0.010 tree 0.038 0.009 0.155 0.789 0.002 0.007 car 0.106 0.019 0.006 0.016 0.837 0.016 clutter 0.340 0.157 0.140 0.031 0.010 0.322 Precision/Correctness 0.881 0.903 0.793 0.863 0.895 0.665 Recall/Completeness 0.899 0.961 0.860 0.789 0.837 0.322 F1 0.890 0.931 0.825 0.824 0.865 0.434 Overall accuracy 0.859	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 0.921 0.029 0.028 0.013 0.000 0.008 building 0.016 0.967 0.010 0.003 0.000 0.004 low_veg 0.035 0.015 0.887 0.054 0.000 0.009 tree 0.029 0.008 0.139 0.816 0.002 0.006 car 0.036 0.020 0.002 0.011 0.916 0.015 clutter 0.345 0.139 0.125 0.028 0.010 0.354 Precision/Correctness 0.903 0.925 0.818 0.889 0.941 0.692 Recall/Completeness 0.921 0.967 0.887 0.816 0.916 0.354 F1 0.912 0.946 0.851 0.851 0.928 0.468 Overall accuracy 0.884
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9061 0.0293 0.0312 0.0108 0.0027 0.0200 building 0.0245 0.9321 0.0239 0.0121 0.0001 0.0072 low_veg 0.0479 0.0219 0.8586 0.0387 0.0002 0.0327 tree 0.0428 0.0149 0.1658 0.7594 0.0013 0.0159 car 0.1198 0.0371 0.0059 0.0093 0.7956 0.0323 clutter 0.1942 0.2419 0.1335 0.0285 0.0243 0.3776 Precision/Correctness 0.867 0.885 0.830 0.911 0.860 0.264 Recall/Completeness 0.906 0.932 0.859 0.759 0.796 0.378 F1 0.886 0.908 0.844 0.828 0.827 0.311 Overall accuracy 0.851 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 0.9311 0.0221 0.0223 0.0064 0.0003 0.0178 building 0.0210 0.9399 0.0219 0.0113 0.0001 0.0058 low_veg 0.0386 0.0152 0.8843 0.0294 0.0001 0.0324 tree 0.0338 0.0129 0.1428 0.7943 0.0009 0.0153 car 0.0562 0.0435 0.0024 0.0046 0.8588 0.0344 clutter 0.1519 0.2496 0.0999 0.0228 0.0291 0.4467 Precision/Correctness 0.897 0.914 0.858 0.934 0.915 0.249 Recall/Completeness 0.931 0.940 0.884 0.794 0.859 0.447 F1 0.914 0.927 0.871 0.859 0.886 0.320 Overall accuracy 0.879 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 0.8950 0.0139 0.0726 0.0099 0.0023 0.0063 building 0.0131 0.9639 0.0085 0.0067 0.0002 0.0075 low_veg 0.0228 0.0087 0.9377 0.0253 0.0001 0.0055 tree 0.0225 0.0069 0.2411 0.7194 0.0008 0.0094 car 0.0962 0.0405 0.0063 0.0138 0.8263 0.0169 clutter 0.1536 0.3154 0.2072 0.0635 0.0255 0.2348 Precision/Correctness 0.876 0.874 0.798 0.953 0.864 0.105 Recall/Completeness 0.895 0.964 0.938 0.719 0.826 0.235 F1 0.886 0.917 0.862 0.820 0.845 0.145 Overall accuracy 0.852 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 0.9242 0.0091 0.0550 0.0059 0.0002 0.0055 building 0.0107 0.9737 0.0052 0.0047 0.0001 0.0057 low_veg 0.0158 0.0059 0.9532 0.0199 0.0001 0.0051 tree 0.0185 0.0059 0.2240 0.7418 0.0005 0.0093 car 0.0380 0.0496 0.0026 0.0092 0.8836 0.0170 clutter 0.1378 0.3538 0.1568 0.0768 0.0288 0.2461 Precision/Correctness 0.904 0.905 0.815 0.965 0.925 0.085 Recall/Completeness 0.924 0.974 0.953 0.742 0.884 0.246 F1 0.914 0.938 0.879 0.839 0.904 0.127 Overall accuracy 0.872 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 0.8657 0.0432 0.0529 0.0183 0.0040 0.0160 building 0.0148 0.9694 0.0064 0.0035 0.0002 0.0057 low_veg 0.0475 0.0455 0.8384 0.0587 0.0002 0.0097 tree 0.0306 0.0127 0.1185 0.8301 0.0025 0.0057 car 0.1166 0.0174 0.0038 0.0206 0.8328 0.0088 clutter 0.3264 0.1977 0.2480 0.0286 0.0127 0.1866 Precision/Correctness 0.874 0.891 0.801 0.905 0.879 0.327 Recall/Completeness 0.866 0.969 0.838 0.830 0.833 0.187 F1 0.870 0.929 0.819 0.866 0.856 0.237 Overall accuracy 0.860 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 0.8965 0.0343 0.0423 0.0118 0.0005 0.0146 building 0.0123 0.9756 0.0047 0.0025 0.0002 0.0047 low_veg 0.0366 0.0330 0.8741 0.0482 0.0000 0.0080 tree 0.0228 0.0107 0.1003 0.8591 0.0020 0.0051 car 0.0409 0.0195 0.0018 0.0145 0.9152 0.0081 clutter 0.3466 0.1922 0.2157 0.0209 0.0152 0.2093 Precision/Correctness 0.904 0.919 0.833 0.931 0.931 0.326 Recall/Completeness 0.896 0.976 0.874 0.859 0.915 0.209 F1 0.900 0.946 0.853 0.894 0.923 0.255 Overall accuracy 0.891 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 0.8629 0.0483 0.0510 0.0186 0.0035 0.0157 building 0.0122 0.9769 0.0055 0.0031 0.0001 0.0023 low_veg 0.0364 0.0252 0.8952 0.0377 0.0001 0.0055 tree 0.0265 0.0123 0.2358 0.7146 0.0012 0.0095 car 0.1035 0.0376 0.0067 0.0282 0.8097 0.0143 clutter 0.2454 0.2354 0.1635 0.0367 0.0067 0.3123 Precision/Correctness 0.866 0.886 0.757 0.919 0.883 0.543 Recall/Completeness 0.863 0.977 0.895 0.715 0.810 0.312 F1 0.865 0.929 0.820 0.804 0.845 0.397 Overall accuracy 0.841 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 0.8943 0.0392 0.0387 0.0126 0.0004 0.0147 building 0.0100 0.9823 0.0039 0.0021 0.0000 0.0017 low_veg 0.0244 0.0168 0.9258 0.0287 0.0000 0.0044 tree 0.0213 0.0109 0.2195 0.7384 0.0010 0.0089 car 0.0377 0.0382 0.0045 0.0198 0.8865 0.0133 clutter 0.2362 0.2266 0.1289 0.0311 0.0071 0.3701 Precision/Correctness 0.899 0.913 0.779 0.943 0.939 0.571 Recall/Completeness 0.894 0.982 0.926 0.738 0.887 0.370 F1 0.897 0.946 0.846 0.828 0.912 0.449 Overall accuracy 0.868 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 0.9044 0.0357 0.0298 0.0239 0.0027 0.0035 building 0.0093 0.9766 0.0071 0.0051 0.0001 0.0017 low_veg 0.0490 0.0311 0.8074 0.1090 0.0002 0.0033 tree 0.0374 0.0054 0.0610 0.8926 0.0015 0.0021 car 0.1143 0.0114 0.0085 0.0264 0.8317 0.0078 clutter 0.4666 0.1922 0.1665 0.0676 0.0072 0.0999 Precision/Correctness 0.848 0.917 0.793 0.793 0.921 0.726 Recall/Completeness 0.904 0.977 0.807 0.893 0.832 0.100 F1 0.875 0.946 0.800 0.840 0.874 0.176 Overall accuracy 0.857 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 0.9300 0.0269 0.0222 0.0177 0.0004 0.0028 building 0.0066 0.9828 0.0052 0.0039 0.0001 0.0014 low_veg 0.0373 0.0193 0.8430 0.0976 0.0000 0.0028 tree 0.0257 0.0034 0.0467 0.9215 0.0010 0.0017 car 0.0359 0.0097 0.0021 0.0174 0.9286 0.0062 clutter 0.4921 0.1722 0.1575 0.0599 0.0079 0.1104 Precision/Correctness 0.872 0.940 0.829 0.829 0.955 0.767 Recall/Completeness 0.930 0.983 0.843 0.922 0.929 0.110 F1 0.900 0.961 0.836 0.873 0.942 0.193 Overall accuracy 0.885 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 0.8674 0.0327 0.0662 0.0148 0.0043 0.0146 building 0.0143 0.9720 0.0046 0.0054 0.0002 0.0035 low_veg 0.0988 0.0221 0.7860 0.0766 0.0005 0.0160 tree 0.0358 0.0086 0.1530 0.7945 0.0020 0.0060 car 0.0978 0.0170 0.0036 0.0259 0.8292 0.0264 clutter 0.5447 0.0600 0.1344 0.0139 0.0022 0.2447 Precision/Correctness 0.714 0.902 0.751 0.860 0.911 0.711 Recall/Completeness 0.867 0.972 0.786 0.794 0.829 0.245 F1 0.783 0.936 0.768 0.826 0.868 0.364 Overall accuracy 0.793 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 0.8889 0.0267 0.0601 0.0102 0.0005 0.0136 building 0.0115 0.9784 0.0030 0.0041 0.0002 0.0028 low_veg 0.0854 0.0152 0.8141 0.0701 0.0002 0.0150 tree 0.0273 0.0070 0.1333 0.8255 0.0016 0.0054 car 0.0368 0.0181 0.0006 0.0183 0.9000 0.0262 clutter 0.5621 0.0550 0.1195 0.0106 0.0018 0.2510 Precision/Correctness 0.727 0.923 0.775 0.884 0.961 0.739 Recall/Completeness 0.889 0.978 0.814 0.826 0.900 0.251 F1 0.800 0.950 0.794 0.854 0.930 0.375 Overall accuracy 0.815 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 0.9112 0.0364 0.0362 0.0050 0.0029 0.0082 building 0.0078 0.9704 0.0158 0.0034 0.0004 0.0022 low_veg 0.0644 0.0390 0.8607 0.0313 0.0004 0.0041 tree 0.0765 0.0088 0.1732 0.7299 0.0058 0.0058 car 0.1141 0.0179 0.0078 0.0055 0.8290 0.0256 clutter 0.3035 0.3307 0.1318 0.0210 0.0108 0.2022 Precision/Correctness 0.901 0.910 0.764 0.916 0.899 0.514 Recall/Completeness 0.911 0.970 0.861 0.730 0.829 0.202 F1 0.906 0.939 0.810 0.812 0.862 0.290 Overall accuracy 0.875 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 0.9320 0.0282 0.0290 0.0031 0.0004 0.0073 building 0.0057 0.9750 0.0144 0.0026 0.0003 0.0018 low_veg 0.0430 0.0276 0.9009 0.0257 0.0000 0.0028 tree 0.0602 0.0068 0.1496 0.7738 0.0044 0.0052 car 0.0383 0.0193 0.0014 0.0021 0.9124 0.0265 clutter 0.3028 0.3442 0.1071 0.0176 0.0109 0.2174 Precision/Correctness 0.930 0.931 0.796 0.939 0.942 0.525 Recall/Completeness 0.932 0.975 0.901 0.774 0.912 0.217 F1 0.931 0.953 0.845 0.848 0.927 0.307 Overall accuracy 0.903 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 0.8921 0.0385 0.0304 0.0315 0.0024 0.0050 building 0.0098 0.9759 0.0076 0.0042 0.0000 0.0024 low_veg 0.0494 0.0228 0.8303 0.0933 0.0002 0.0039 tree 0.0446 0.0076 0.1183 0.8245 0.0024 0.0026 car 0.1282 0.0141 0.0016 0.0076 0.8385 0.0100 clutter 0.2416 0.3585 0.1505 0.0358 0.0279 0.1857 Precision/Correctness 0.912 0.903 0.833 0.767 0.914 0.617 Recall/Completeness 0.892 0.976 0.830 0.824 0.839 0.186 F1 0.902 0.938 0.832 0.795 0.875 0.286 Overall accuracy 0.871 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 0.9155 0.0301 0.0252 0.0251 0.0002 0.0038 building 0.0076 0.9803 0.0066 0.0034 0.0000 0.0021 low_veg 0.0440 0.0164 0.8530 0.0836 0.0001 0.0029 tree 0.0315 0.0057 0.1047 0.8539 0.0019 0.0024 car 0.0454 0.0156 0.0004 0.0049 0.9249 0.0088 clutter 0.2146 0.3623 0.1390 0.0361 0.0350 0.2131 Precision/Correctness 0.936 0.926 0.860 0.797 0.943 0.645 Recall/Completeness 0.916 0.980 0.853 0.854 0.925 0.213 F1 0.926 0.952 0.856 0.825 0.934 0.320 Overall accuracy 0.898 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 0.9127 0.0191 0.0314 0.0257 0.0023 0.0087 building 0.0166 0.9762 0.0039 0.0021 0.0001 0.0011 low_veg 0.0620 0.0196 0.8194 0.0847 0.0005 0.0139 tree 0.0473 0.0092 0.0972 0.8337 0.0038 0.0089 car 0.1123 0.0102 0.0122 0.0146 0.8229 0.0278 clutter 0.2168 0.2300 0.1203 0.0661 0.0093 0.3575 Precision/Correctness 0.849 0.897 0.827 0.829 0.884 0.777 Recall/Completeness 0.913 0.976 0.819 0.834 0.823 0.357 F1 0.880 0.935 0.823 0.831 0.853 0.490 Overall accuracy 0.852 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 0.9348 0.0144 0.0236 0.0199 0.0003 0.0071 building 0.0150 0.9798 0.0027 0.0017 0.0001 0.0009 low_veg 0.0470 0.0137 0.8472 0.0789 0.0001 0.0130 tree 0.0373 0.0077 0.0875 0.8567 0.0030 0.0077 car 0.0324 0.0105 0.0086 0.0092 0.9123 0.0271 clutter 0.2080 0.1978 0.1120 0.0676 0.0095 0.4052 Precision/Correctness 0.877 0.923 0.851 0.852 0.919 0.807 Recall/Completeness 0.935 0.980 0.847 0.857 0.912 0.405 F1 0.905 0.950 0.849 0.854 0.916 0.539 Overall accuracy 0.878 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 0.8907 0.0309 0.0459 0.0189 0.0023 0.0114 building 0.0167 0.9600 0.0130 0.0061 0.0002 0.0040 low_veg 0.0473 0.0178 0.8118 0.1085 0.0001 0.0145 tree 0.0337 0.0074 0.1030 0.8513 0.0020 0.0026 car 0.1313 0.0121 0.0078 0.0148 0.8220 0.0120 clutter 0.3238 0.1542 0.1186 0.0185 0.0068 0.3780 Precision/Correctness 0.897 0.898 0.734 0.790 0.927 0.763 Recall/Completeness 0.891 0.960 0.812 0.851 0.822 0.378 F1 0.894 0.928 0.771 0.820 0.872 0.505 Overall accuracy 0.854 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 0.9087 0.0257 0.0399 0.0146 0.0005 0.0106 building 0.0125 0.9677 0.0116 0.0047 0.0001 0.0034 low_veg 0.0282 0.0115 0.8428 0.1033 0.0001 0.0142 tree 0.0239 0.0052 0.0907 0.8763 0.0016 0.0022 car 0.0537 0.0121 0.0018 0.0119 0.9088 0.0117 clutter 0.3326 0.1137 0.1111 0.0165 0.0071 0.4189 Precision/Correctness 0.918 0.925 0.759 0.818 0.956 0.783 Recall/Completeness 0.909 0.968 0.843 0.876 0.909 0.419 F1 0.913 0.946 0.798 0.846 0.932 0.546 Overall accuracy 0.879 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 0.9186 0.0376 0.0242 0.0156 0.0014 0.0028 building 0.0246 0.9541 0.0160 0.0017 0.0002 0.0033 low_veg 0.0192 0.0093 0.9334 0.0372 0.0000 0.0010 tree 0.0729 0.0034 0.0969 0.8233 0.0015 0.0021 car 0.1027 0.0547 0.0026 0.0033 0.8039 0.0328 clutter 0.5861 0.0712 0.1471 0.0431 0.0090 0.1434 Precision/Correctness 0.914 0.924 0.889 0.860 0.913 0.519 Recall/Completeness 0.919 0.954 0.933 0.823 0.804 0.143 F1 0.916 0.939 0.911 0.841 0.855 0.225 Overall accuracy 0.902 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 0.9374 0.0327 0.0174 0.0103 0.0002 0.0019 building 0.0217 0.9603 0.0139 0.0012 0.0002 0.0028 low_veg 0.0119 0.0046 0.9519 0.0309 0.0000 0.0008 tree 0.0551 0.0021 0.0846 0.8553 0.0012 0.0017 car 0.0277 0.0649 0.0025 0.0018 0.8762 0.0268 clutter 0.6179 0.0654 0.1207 0.0413 0.0084 0.1463 Precision/Correctness 0.931 0.939 0.911 0.890 0.955 0.564 Recall/Completeness 0.937 0.960 0.952 0.855 0.876 0.146 F1 0.934 0.949 0.931 0.872 0.914 0.232 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: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9326 0.0229 0.0245 0.0118 0.0056 0.0027 building 0.0433 0.9459 0.0034 0.0016 0.0006 0.0052 low_veg 0.0563 0.0211 0.8435 0.0766 0.0006 0.0021 tree 0.0431 0.0081 0.0970 0.8452 0.0042 0.0024 car 0.0846 0.0094 0.0038 0.0128 0.8812 0.0082 clutter 0.3315 0.2724 0.2317 0.0368 0.0195 0.1081 Precision/Correctness 0.910 0.936 0.837 0.815 0.893 0.423 Recall/Completeness 0.933 0.946 0.843 0.845 0.881 0.108 F1 0.921 0.941 0.840 0.830 0.887 0.172 Overall accuracy 0.894 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 0.9531 0.0169 0.0185 0.0087 0.0007 0.0022 building 0.0404 0.9517 0.0018 0.0010 0.0004 0.0047 low_veg 0.0405 0.0122 0.8763 0.0698 0.0001 0.0012 tree 0.0327 0.0062 0.0844 0.8712 0.0032 0.0023 car 0.0233 0.0083 0.0013 0.0084 0.9505 0.0082 clutter 0.3215 0.2947 0.2241 0.0330 0.0188 0.1079 Precision/Correctness 0.928 0.951 0.864 0.845 0.956 0.432 Recall/Completeness 0.953 0.952 0.876 0.871 0.951 0.108 F1 0.940 0.951 0.870 0.858 0.953 0.173 Overall accuracy 0.916 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 0.9002 0.0181 0.0453 0.0145 0.0058 0.0160 building 0.0126 0.9290 0.0501 0.0055 0.0001 0.0026 low_veg 0.0370 0.0059 0.8856 0.0602 0.0009 0.0104 tree 0.0277 0.0059 0.1747 0.7848 0.0034 0.0034 car 0.0817 0.0124 0.0118 0.0168 0.8690 0.0084 clutter 0.1777 0.0752 0.1958 0.0236 0.0175 0.5103 Precision/Correctness 0.929 0.941 0.659 0.874 0.871 0.773 Recall/Completeness 0.900 0.929 0.886 0.785 0.869 0.510 F1 0.915 0.935 0.756 0.827 0.870 0.615 Overall accuracy 0.863 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 0.9242 0.0136 0.0355 0.0106 0.0010 0.0151 building 0.0090 0.9356 0.0489 0.0045 0.0001 0.0019 low_veg 0.0217 0.0025 0.9142 0.0527 0.0001 0.0088 tree 0.0197 0.0044 0.1557 0.8145 0.0026 0.0030 car 0.0225 0.0124 0.0030 0.0116 0.9428 0.0076 clutter 0.1717 0.0742 0.1833 0.0210 0.0173 0.5326 Precision/Correctness 0.948 0.954 0.679 0.900 0.926 0.797 Recall/Completeness 0.924 0.936 0.914 0.814 0.943 0.533 F1 0.936 0.945 0.779 0.855 0.934 0.638 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: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8920 0.0614 0.0108 0.0269 0.0020 0.0069 building 0.0340 0.9457 0.0025 0.0020 0.0001 0.0157 low_veg 0.0494 0.0314 0.7561 0.1606 0.0004 0.0020 tree 0.0187 0.0131 0.1002 0.8662 0.0006 0.0013 car 0.0839 0.0375 0.0033 0.0231 0.8444 0.0078 clutter 0.3056 0.0346 0.0561 0.0048 0.0070 0.5919 Precision/Correctness 0.912 0.847 0.802 0.687 0.870 0.879 Recall/Completeness 0.892 0.946 0.756 0.866 0.844 0.592 F1 0.902 0.894 0.779 0.766 0.857 0.707 Overall accuracy 0.861 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 0.9085 0.0577 0.0074 0.0196 0.0004 0.0064 building 0.0326 0.9489 0.0020 0.0012 0.0001 0.0153 low_veg 0.0386 0.0263 0.7836 0.1499 0.0001 0.0015 tree 0.0109 0.0110 0.0921 0.8846 0.0004 0.0010 car 0.0225 0.0428 0.0015 0.0153 0.9106 0.0073 clutter 0.2987 0.0324 0.0543 0.0040 0.0063 0.6042 Precision/Correctness 0.920 0.859 0.825 0.721 0.922 0.888 Recall/Completeness 0.908 0.949 0.784 0.885 0.911 0.604 F1 0.914 0.902 0.804 0.795 0.916 0.719 Overall accuracy 0.876 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image