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 0.916 0.015 0.039 0.016 0.003 0.012 building 0.016 0.968 0.007 0.003 0.000 0.006 low_veg 0.044 0.009 0.876 0.058 0.000 0.013 tree 0.038 0.006 0.136 0.815 0.002 0.003 car 0.077 0.006 0.004 0.013 0.886 0.013 clutter 0.300 0.104 0.162 0.022 0.011 0.401 Precision/Correctness 0.893 0.948 0.806 0.879 0.902 0.682 Recall/Completeness 0.916 0.968 0.876 0.815 0.886 0.401 F1 0.904 0.958 0.840 0.846 0.894 0.505 Overall accuracy 0.879	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 0.935 0.011 0.032 0.011 0.000 0.011 building 0.012 0.976 0.004 0.002 0.000 0.006 low_veg 0.032 0.005 0.902 0.050 0.000 0.011 tree 0.028 0.004 0.120 0.842 0.002 0.002 car 0.014 0.007 0.001 0.009 0.955 0.014 clutter 0.300 0.089 0.148 0.018 0.011 0.433 Precision/Correctness 0.915 0.964 0.831 0.905 0.943 0.704 Recall/Completeness 0.935 0.976 0.902 0.842 0.955 0.433 F1 0.925 0.970 0.865 0.872 0.949 0.536 Overall accuracy 0.902
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9003 0.0174 0.0423 0.0105 0.0026 0.0270 building 0.0102 0.9685 0.0085 0.0075 0.0001 0.0052 low_veg 0.0312 0.0084 0.8846 0.0399 0.0002 0.0357 tree 0.0392 0.0107 0.1421 0.8009 0.0012 0.0060 car 0.0872 0.0077 0.0027 0.0084 0.8502 0.0439 clutter 0.1390 0.1897 0.1456 0.0369 0.0162 0.4726 Precision/Correctness 0.902 0.932 0.849 0.917 0.883 0.309 Recall/Completeness 0.900 0.968 0.885 0.801 0.850 0.473 F1 0.901 0.950 0.866 0.855 0.866 0.374 Overall accuracy 0.876 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 0.9237 0.0119 0.0325 0.0064 0.0003 0.0252 building 0.0061 0.9789 0.0044 0.0063 0.0001 0.0043 low_veg 0.0230 0.0045 0.9066 0.0297 0.0001 0.0361 tree 0.0298 0.0089 0.1215 0.8337 0.0008 0.0052 car 0.0305 0.0086 0.0003 0.0040 0.9048 0.0518 clutter 0.0999 0.1926 0.1013 0.0328 0.0184 0.5550 Precision/Correctness 0.931 0.953 0.877 0.941 0.938 0.287 Recall/Completeness 0.924 0.979 0.907 0.834 0.905 0.555 F1 0.927 0.966 0.891 0.884 0.921 0.379 Overall accuracy 0.903 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 0.9002 0.0063 0.0770 0.0100 0.0016 0.0048 building 0.0071 0.9745 0.0103 0.0055 0.0000 0.0026 low_veg 0.0171 0.0028 0.9394 0.0294 0.0000 0.0113 tree 0.0189 0.0033 0.2090 0.7673 0.0007 0.0008 car 0.0697 0.0114 0.0031 0.0081 0.9035 0.0043 clutter 0.0908 0.2950 0.1827 0.0768 0.0056 0.3492 Precision/Correctness 0.902 0.939 0.816 0.950 0.908 0.175 Recall/Completeness 0.900 0.975 0.939 0.767 0.904 0.349 F1 0.901 0.956 0.873 0.849 0.906 0.233 Overall accuracy 0.872 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 0.9296 0.0030 0.0574 0.0059 0.0001 0.0040 building 0.0039 0.9857 0.0049 0.0033 0.0000 0.0022 low_veg 0.0123 0.0011 0.9522 0.0229 0.0000 0.0114 tree 0.0146 0.0024 0.1923 0.7895 0.0005 0.0007 car 0.0122 0.0116 0.0001 0.0043 0.9668 0.0050 clutter 0.0592 0.3299 0.1183 0.0923 0.0068 0.3936 Precision/Correctness 0.927 0.962 0.834 0.963 0.954 0.149 Recall/Completeness 0.930 0.986 0.952 0.789 0.967 0.394 F1 0.928 0.974 0.889 0.868 0.960 0.216 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_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9023 0.0183 0.0445 0.0162 0.0037 0.0151 building 0.0112 0.9751 0.0051 0.0032 0.0000 0.0054 low_veg 0.0548 0.0175 0.8676 0.0518 0.0002 0.0080 tree 0.0339 0.0077 0.1090 0.8448 0.0024 0.0022 car 0.0729 0.0081 0.0041 0.0157 0.8957 0.0035 clutter 0.2964 0.1255 0.2549 0.0236 0.0090 0.2906 Precision/Correctness 0.878 0.947 0.821 0.917 0.898 0.479 Recall/Completeness 0.902 0.975 0.868 0.845 0.896 0.291 F1 0.890 0.961 0.844 0.879 0.897 0.362 Overall accuracy 0.884 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 0.9305 0.0116 0.0339 0.0099 0.0003 0.0138 building 0.0082 0.9835 0.0017 0.0022 0.0000 0.0043 low_veg 0.0432 0.0088 0.8998 0.0421 0.0001 0.0060 tree 0.0256 0.0061 0.0908 0.8741 0.0018 0.0016 car 0.0094 0.0091 0.0018 0.0097 0.9662 0.0037 clutter 0.3102 0.1164 0.2215 0.0151 0.0113 0.3254 Precision/Correctness 0.907 0.967 0.855 0.941 0.948 0.486 Recall/Completeness 0.930 0.984 0.900 0.874 0.966 0.325 F1 0.919 0.975 0.877 0.906 0.957 0.390 Overall accuracy 0.913 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 0.8836 0.0171 0.0644 0.0182 0.0036 0.0131 building 0.0114 0.9765 0.0059 0.0025 0.0001 0.0036 low_veg 0.0309 0.0091 0.9078 0.0441 0.0001 0.0080 tree 0.0250 0.0048 0.1960 0.7691 0.0013 0.0038 car 0.0644 0.0074 0.0031 0.0274 0.8813 0.0164 clutter 0.2080 0.1507 0.1860 0.0265 0.0106 0.4182 Precision/Correctness 0.884 0.947 0.776 0.919 0.879 0.647 Recall/Completeness 0.884 0.977 0.908 0.769 0.881 0.418 F1 0.884 0.962 0.837 0.837 0.880 0.508 Overall accuracy 0.867 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 0.9145 0.0111 0.0516 0.0116 0.0003 0.0110 building 0.0089 0.9844 0.0024 0.0016 0.0001 0.0027 low_veg 0.0205 0.0040 0.9348 0.0343 0.0000 0.0063 tree 0.0196 0.0036 0.1797 0.7930 0.0011 0.0031 car 0.0096 0.0083 0.0008 0.0172 0.9464 0.0178 clutter 0.1993 0.1446 0.1549 0.0200 0.0126 0.4687 Precision/Correctness 0.914 0.966 0.800 0.943 0.929 0.682 Recall/Completeness 0.914 0.984 0.935 0.793 0.946 0.469 F1 0.914 0.975 0.862 0.862 0.938 0.556 Overall accuracy 0.893 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 0.9299 0.0090 0.0322 0.0226 0.0025 0.0038 building 0.0112 0.9743 0.0067 0.0038 0.0001 0.0039 low_veg 0.0507 0.0095 0.8294 0.1050 0.0002 0.0053 tree 0.0314 0.0034 0.0499 0.9124 0.0022 0.0006 car 0.0723 0.0008 0.0065 0.0228 0.8926 0.0050 clutter 0.4900 0.1141 0.1750 0.0685 0.0104 0.1421 Precision/Correctness 0.850 0.963 0.799 0.804 0.915 0.738 Recall/Completeness 0.930 0.974 0.829 0.912 0.893 0.142 F1 0.888 0.969 0.814 0.855 0.904 0.238 Overall accuracy 0.874 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 0.9497 0.0053 0.0258 0.0160 0.0004 0.0027 building 0.0061 0.9851 0.0027 0.0026 0.0001 0.0033 low_veg 0.0375 0.0044 0.8604 0.0941 0.0000 0.0035 tree 0.0196 0.0019 0.0370 0.9397 0.0016 0.0003 car 0.0074 0.0002 0.0009 0.0143 0.9725 0.0047 clutter 0.5122 0.0991 0.1658 0.0598 0.0115 0.1516 Precision/Correctness 0.875 0.976 0.835 0.840 0.943 0.780 Recall/Completeness 0.950 0.985 0.860 0.940 0.972 0.152 F1 0.911 0.981 0.847 0.887 0.957 0.254 Overall accuracy 0.900 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 0.8873 0.0178 0.0661 0.0128 0.0042 0.0119 building 0.0092 0.9772 0.0044 0.0039 0.0000 0.0053 low_veg 0.0853 0.0139 0.8170 0.0653 0.0004 0.0182 tree 0.0362 0.0061 0.1324 0.8173 0.0024 0.0056 car 0.0692 0.0104 0.0048 0.0220 0.8766 0.0170 clutter 0.4535 0.0439 0.1402 0.0116 0.0016 0.3491 Precision/Correctness 0.751 0.936 0.768 0.882 0.920 0.780 Recall/Completeness 0.887 0.977 0.817 0.817 0.877 0.349 F1 0.814 0.956 0.792 0.848 0.898 0.482 Overall accuracy 0.822 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 0.9078 0.0130 0.0601 0.0083 0.0003 0.0105 building 0.0062 0.9845 0.0022 0.0029 0.0000 0.0042 low_veg 0.0730 0.0083 0.8447 0.0575 0.0000 0.0165 tree 0.0270 0.0045 0.1139 0.8480 0.0018 0.0048 car 0.0203 0.0110 0.0013 0.0157 0.9337 0.0181 clutter 0.4662 0.0401 0.1270 0.0088 0.0012 0.3568 Precision/Correctness 0.766 0.953 0.791 0.906 0.970 0.809 Recall/Completeness 0.908 0.985 0.845 0.848 0.934 0.357 F1 0.831 0.969 0.817 0.876 0.951 0.495 Overall accuracy 0.843 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 0.9073 0.0174 0.0582 0.0056 0.0030 0.0085 building 0.0058 0.9851 0.0044 0.0030 0.0001 0.0017 low_veg 0.0560 0.0199 0.8868 0.0302 0.0005 0.0065 tree 0.0742 0.0105 0.1422 0.7597 0.0071 0.0064 car 0.0860 0.0037 0.0073 0.0049 0.8747 0.0234 clutter 0.3196 0.2210 0.1669 0.0117 0.0048 0.2761 Precision/Correctness 0.907 0.947 0.766 0.922 0.905 0.576 Recall/Completeness 0.907 0.985 0.887 0.760 0.875 0.276 F1 0.907 0.966 0.822 0.833 0.890 0.373 Overall accuracy 0.890 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 0.9253 0.0110 0.0528 0.0035 0.0005 0.0069 building 0.0035 0.9908 0.0022 0.0023 0.0000 0.0012 low_veg 0.0353 0.0106 0.9267 0.0232 0.0000 0.0042 tree 0.0571 0.0086 0.1207 0.8027 0.0054 0.0055 car 0.0159 0.0032 0.0017 0.0016 0.9484 0.0292 clutter 0.3314 0.2234 0.1493 0.0087 0.0035 0.2838 Precision/Correctness 0.935 0.964 0.793 0.946 0.949 0.593 Recall/Completeness 0.925 0.991 0.927 0.803 0.948 0.284 F1 0.930 0.977 0.855 0.869 0.949 0.384 Overall accuracy 0.916 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 0.9287 0.0144 0.0277 0.0203 0.0023 0.0067 building 0.0093 0.9806 0.0051 0.0027 0.0000 0.0024 low_veg 0.0609 0.0094 0.8556 0.0668 0.0001 0.0073 tree 0.0522 0.0060 0.1147 0.8228 0.0024 0.0019 car 0.0944 0.0044 0.0011 0.0081 0.8855 0.0066 clutter 0.2521 0.2377 0.1649 0.0147 0.0538 0.2768 Precision/Correctness 0.908 0.949 0.844 0.829 0.889 0.655 Recall/Completeness 0.929 0.981 0.856 0.823 0.886 0.277 F1 0.918 0.964 0.850 0.826 0.887 0.389 Overall accuracy 0.894 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 0.9476 0.0090 0.0232 0.0149 0.0002 0.0051 building 0.0052 0.9876 0.0035 0.0020 0.0000 0.0018 low_veg 0.0535 0.0048 0.8784 0.0579 0.0000 0.0055 tree 0.0361 0.0042 0.1006 0.8556 0.0020 0.0015 car 0.0185 0.0048 0.0001 0.0055 0.9641 0.0070 clutter 0.2171 0.2292 0.1634 0.0127 0.0677 0.3099 Precision/Correctness 0.935 0.966 0.870 0.859 0.912 0.682 Recall/Completeness 0.948 0.988 0.878 0.856 0.964 0.310 F1 0.941 0.977 0.874 0.857 0.937 0.426 Overall accuracy 0.919 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 0.9297 0.0076 0.0267 0.0244 0.0019 0.0097 building 0.0169 0.9746 0.0045 0.0016 0.0001 0.0022 low_veg 0.0578 0.0076 0.8414 0.0831 0.0003 0.0097 tree 0.0515 0.0063 0.0986 0.8358 0.0035 0.0043 car 0.0916 0.0124 0.0081 0.0144 0.8585 0.0150 clutter 0.2084 0.1769 0.1835 0.0368 0.0084 0.3860 Precision/Correctness 0.855 0.931 0.816 0.843 0.901 0.822 Recall/Completeness 0.930 0.975 0.841 0.836 0.858 0.386 F1 0.891 0.952 0.828 0.840 0.879 0.525 Overall accuracy 0.864 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 0.9500 0.0044 0.0186 0.0187 0.0001 0.0082 building 0.0146 0.9802 0.0021 0.0011 0.0001 0.0019 low_veg 0.0433 0.0042 0.8668 0.0778 0.0001 0.0078 tree 0.0406 0.0051 0.0891 0.8588 0.0026 0.0038 car 0.0137 0.0143 0.0039 0.0083 0.9434 0.0165 clutter 0.1970 0.1507 0.1754 0.0347 0.0086 0.4336 Precision/Correctness 0.883 0.951 0.842 0.866 0.931 0.849 Recall/Completeness 0.950 0.980 0.867 0.859 0.943 0.434 F1 0.915 0.965 0.854 0.863 0.937 0.574 Overall accuracy 0.889 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 0.9015 0.0086 0.0505 0.0175 0.0020 0.0199 building 0.0127 0.9674 0.0085 0.0048 0.0003 0.0063 low_veg 0.0404 0.0065 0.8193 0.1159 0.0001 0.0178 tree 0.0375 0.0034 0.0864 0.8686 0.0022 0.0018 car 0.1058 0.0064 0.0051 0.0120 0.8590 0.0118 clutter 0.2856 0.1188 0.1354 0.0156 0.0093 0.4352 Precision/Correctness 0.909 0.947 0.737 0.794 0.926 0.708 Recall/Completeness 0.901 0.967 0.819 0.869 0.859 0.435 F1 0.905 0.957 0.776 0.830 0.891 0.539 Overall accuracy 0.868 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 0.9175 0.0045 0.0449 0.0131 0.0005 0.0195 building 0.0083 0.9756 0.0066 0.0036 0.0002 0.0057 low_veg 0.0242 0.0024 0.8462 0.1100 0.0000 0.0170 tree 0.0266 0.0017 0.0754 0.8931 0.0016 0.0017 car 0.0223 0.0067 0.0008 0.0099 0.9478 0.0125 clutter 0.2856 0.0867 0.1239 0.0138 0.0098 0.4802 Precision/Correctness 0.930 0.968 0.761 0.822 0.949 0.722 Recall/Completeness 0.917 0.976 0.846 0.893 0.948 0.480 F1 0.924 0.972 0.801 0.856 0.948 0.577 Overall accuracy 0.892 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 0.9249 0.0221 0.0242 0.0150 0.0016 0.0123 building 0.0287 0.9478 0.0171 0.0014 0.0002 0.0049 low_veg 0.0205 0.0042 0.9421 0.0310 0.0000 0.0022 tree 0.0693 0.0028 0.0941 0.8308 0.0012 0.0018 car 0.0616 0.0148 0.0002 0.0029 0.9033 0.0171 clutter 0.3862 0.0433 0.3157 0.0183 0.0025 0.2340 Precision/Correctness 0.924 0.955 0.878 0.878 0.931 0.416 Recall/Completeness 0.925 0.948 0.942 0.831 0.903 0.234 F1 0.924 0.951 0.909 0.854 0.917 0.300 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: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9415 0.0191 0.0188 0.0096 0.0002 0.0109 building 0.0246 0.9582 0.0121 0.0008 0.0002 0.0041 low_veg 0.0125 0.0020 0.9583 0.0257 0.0000 0.0015 tree 0.0508 0.0017 0.0816 0.8636 0.0008 0.0016 car 0.0065 0.0181 0.0000 0.0008 0.9573 0.0172 clutter 0.3925 0.0416 0.3200 0.0172 0.0018 0.2269 Precision/Correctness 0.941 0.964 0.899 0.907 0.977 0.415 Recall/Completeness 0.942 0.958 0.958 0.864 0.957 0.227 F1 0.941 0.961 0.928 0.885 0.967 0.293 Overall accuracy 0.928 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 0.9536 0.0063 0.0228 0.0096 0.0045 0.0032 building 0.0419 0.9408 0.0050 0.0010 0.0003 0.0109 low_veg 0.0550 0.0061 0.8674 0.0645 0.0005 0.0065 tree 0.0439 0.0075 0.0882 0.8539 0.0053 0.0013 car 0.0636 0.0026 0.0023 0.0102 0.9147 0.0066 clutter 0.2987 0.1430 0.2672 0.0210 0.0156 0.2545 Precision/Correctness 0.916 0.974 0.842 0.846 0.912 0.501 Recall/Completeness 0.954 0.941 0.867 0.854 0.915 0.254 F1 0.934 0.957 0.855 0.850 0.913 0.337 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_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9702 0.0028 0.0175 0.0067 0.0003 0.0026 building 0.0374 0.9491 0.0023 0.0006 0.0001 0.0105 low_veg 0.0392 0.0018 0.8958 0.0584 0.0001 0.0047 tree 0.0328 0.0059 0.0749 0.8811 0.0041 0.0011 car 0.0074 0.0018 0.0004 0.0059 0.9776 0.0068 clutter 0.2850 0.1578 0.2622 0.0167 0.0149 0.2634 Precision/Correctness 0.934 0.983 0.869 0.874 0.966 0.508 Recall/Completeness 0.970 0.949 0.896 0.881 0.978 0.263 F1 0.952 0.966 0.882 0.877 0.972 0.347 Overall accuracy 0.929 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 0.9193 0.0056 0.0465 0.0099 0.0051 0.0137 building 0.0090 0.9694 0.0103 0.0029 0.0005 0.0079 low_veg 0.0425 0.0050 0.8749 0.0562 0.0011 0.0202 tree 0.0392 0.0049 0.1470 0.8018 0.0049 0.0021 car 0.0666 0.0025 0.0061 0.0098 0.9062 0.0088 clutter 0.1665 0.0126 0.1969 0.0082 0.0147 0.6011 Precision/Correctness 0.929 0.981 0.704 0.902 0.878 0.773 Recall/Completeness 0.919 0.969 0.875 0.802 0.906 0.601 F1 0.924 0.975 0.780 0.849 0.892 0.676 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 0.9389 0.0026 0.0391 0.0065 0.0005 0.0124 building 0.0046 0.9788 0.0071 0.0021 0.0004 0.0070 low_veg 0.0269 0.0027 0.9042 0.0470 0.0001 0.0191 tree 0.0292 0.0033 0.1307 0.8312 0.0040 0.0016 car 0.0127 0.0024 0.0010 0.0061 0.9677 0.0102 clutter 0.1600 0.0111 0.1859 0.0067 0.0149 0.6214 Precision/Correctness 0.949 0.989 0.725 0.928 0.930 0.794 Recall/Completeness 0.939 0.979 0.904 0.831 0.968 0.621 F1 0.944 0.984 0.805 0.877 0.949 0.697 Overall accuracy 0.909 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 0.9152 0.0342 0.0094 0.0242 0.0017 0.0152 building 0.0240 0.9482 0.0011 0.0021 0.0002 0.0243 low_veg 0.0575 0.0151 0.7733 0.1467 0.0002 0.0071 tree 0.0242 0.0064 0.0927 0.8753 0.0006 0.0009 car 0.0555 0.0139 0.0016 0.0265 0.8877 0.0148 clutter 0.2289 0.0217 0.0643 0.0021 0.0087 0.6743 Precision/Correctness 0.929 0.909 0.813 0.710 0.873 0.813 Recall/Completeness 0.915 0.948 0.773 0.875 0.888 0.674 F1 0.922 0.928 0.793 0.784 0.880 0.737 Overall accuracy 0.883 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 0.9302 0.0306 0.0070 0.0172 0.0003 0.0146 building 0.0215 0.9528 0.0005 0.0012 0.0002 0.0238 low_veg 0.0443 0.0109 0.8008 0.1378 0.0000 0.0062 tree 0.0147 0.0046 0.0846 0.8950 0.0003 0.0009 car 0.0056 0.0161 0.0006 0.0180 0.9437 0.0161 clutter 0.2213 0.0199 0.0632 0.0016 0.0081 0.6860 Precision/Correctness 0.938 0.921 0.832 0.745 0.914 0.823 Recall/Completeness 0.930 0.953 0.801 0.895 0.944 0.686 F1 0.934 0.937 0.816 0.813 0.929 0.748 Overall accuracy 0.898 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image