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 0.907 0.037 0.035 0.016 0.002 0.003 building 0.024 0.961 0.005 0.008 0.000 0.001 low_veg 0.068 0.037 0.823 0.068 0.000 0.003 tree 0.072 0.013 0.178 0.734 0.003 0.002 car 0.155 0.019 0.005 0.016 0.802 0.003 clutter 0.444 0.157 0.140 0.046 0.009 0.205 Precision/Correctness 0.837 0.887 0.779 0.840 0.902 0.816 Recall/Completeness 0.907 0.961 0.823 0.734 0.802 0.205 F1 0.870 0.923 0.800 0.783 0.849 0.328 Overall accuracy 0.838	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 0.927 0.030 0.028 0.012 0.000 0.002 building 0.021 0.967 0.004 0.007 0.000 0.001 low_veg 0.056 0.029 0.852 0.060 0.000 0.003 tree 0.062 0.011 0.163 0.761 0.002 0.002 car 0.072 0.020 0.002 0.011 0.892 0.003 clutter 0.457 0.136 0.128 0.043 0.008 0.228 Precision/Correctness 0.859 0.911 0.802 0.867 0.941 0.839 Recall/Completeness 0.927 0.967 0.852 0.761 0.892 0.228 F1 0.892 0.938 0.826 0.810 0.916 0.358 Overall accuracy 0.863
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9245 0.0313 0.0242 0.0113 0.0023 0.0064 building 0.0236 0.9277 0.0136 0.0333 0.0002 0.0016 low_veg 0.1051 0.0298 0.8058 0.0512 0.0002 0.0079 tree 0.0819 0.0159 0.1953 0.7018 0.0016 0.0035 car 0.1726 0.0347 0.0041 0.0086 0.7752 0.0048 clutter 0.3559 0.2254 0.1083 0.0439 0.0322 0.2343 Precision/Correctness 0.780 0.872 0.814 0.865 0.846 0.468 Recall/Completeness 0.925 0.928 0.806 0.702 0.775 0.234 F1 0.846 0.899 0.810 0.775 0.809 0.312 Overall accuracy 0.822 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 0.9464 0.0234 0.0167 0.0073 0.0003 0.0059 building 0.0190 0.9351 0.0124 0.0322 0.0001 0.0011 low_veg 0.0969 0.0223 0.8318 0.0416 0.0001 0.0073 tree 0.0730 0.0139 0.1751 0.7337 0.0012 0.0031 car 0.0950 0.0394 0.0019 0.0056 0.8524 0.0058 clutter 0.3362 0.2267 0.0832 0.0338 0.0362 0.2839 Precision/Correctness 0.806 0.903 0.838 0.890 0.897 0.466 Recall/Completeness 0.946 0.935 0.832 0.734 0.852 0.284 F1 0.871 0.919 0.835 0.804 0.874 0.353 Overall accuracy 0.849 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 0.9100 0.0143 0.0618 0.0113 0.0020 0.0007 building 0.0127 0.9674 0.0076 0.0110 0.0004 0.0008 low_veg 0.0351 0.0141 0.9126 0.0359 0.0001 0.0021 tree 0.0454 0.0114 0.2231 0.7183 0.0008 0.0009 car 0.1527 0.0304 0.0051 0.0100 0.7982 0.0037 clutter 0.3093 0.2416 0.1482 0.0794 0.0355 0.1860 Precision/Correctness 0.807 0.835 0.807 0.935 0.857 0.328 Recall/Completeness 0.910 0.967 0.913 0.718 0.798 0.186 F1 0.855 0.897 0.857 0.813 0.826 0.237 Overall accuracy 0.844 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 0.9360 0.0094 0.0469 0.0070 0.0002 0.0005 building 0.0093 0.9763 0.0053 0.0084 0.0003 0.0005 low_veg 0.0273 0.0112 0.9304 0.0291 0.0001 0.0021 tree 0.0410 0.0105 0.2078 0.7392 0.0006 0.0008 car 0.0850 0.0366 0.0025 0.0057 0.8664 0.0037 clutter 0.2814 0.2595 0.1109 0.0955 0.0375 0.2153 Precision/Correctness 0.833 0.864 0.824 0.950 0.911 0.303 Recall/Completeness 0.936 0.976 0.930 0.739 0.866 0.215 F1 0.882 0.917 0.874 0.832 0.888 0.252 Overall accuracy 0.863 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 0.8820 0.0484 0.0413 0.0223 0.0032 0.0028 building 0.0172 0.9736 0.0013 0.0066 0.0002 0.0012 low_veg 0.0619 0.0696 0.7837 0.0820 0.0002 0.0025 tree 0.0497 0.0148 0.1211 0.8107 0.0029 0.0008 car 0.1580 0.0197 0.0022 0.0227 0.7965 0.0009 clutter 0.3961 0.2395 0.2007 0.0552 0.0142 0.0943 Precision/Correctness 0.840 0.864 0.806 0.870 0.878 0.560 Recall/Completeness 0.882 0.974 0.784 0.811 0.796 0.094 F1 0.861 0.916 0.795 0.840 0.835 0.161 Overall accuracy 0.845 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 0.9107 0.0388 0.0318 0.0157 0.0006 0.0025 building 0.0145 0.9789 0.0003 0.0052 0.0001 0.0008 low_veg 0.0499 0.0560 0.8215 0.0707 0.0001 0.0018 tree 0.0416 0.0127 0.1055 0.8372 0.0024 0.0006 car 0.0690 0.0215 0.0009 0.0167 0.8914 0.0005 clutter 0.4217 0.2346 0.1697 0.0491 0.0173 0.1077 Precision/Correctness 0.871 0.895 0.837 0.899 0.919 0.585 Recall/Completeness 0.911 0.979 0.821 0.837 0.891 0.108 F1 0.891 0.935 0.829 0.867 0.905 0.182 Overall accuracy 0.876 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 0.8821 0.0477 0.0483 0.0168 0.0031 0.0019 building 0.0130 0.9800 0.0013 0.0052 0.0001 0.0004 low_veg 0.0543 0.0408 0.8515 0.0522 0.0001 0.0011 tree 0.0521 0.0153 0.2352 0.6936 0.0016 0.0023 car 0.1572 0.0202 0.0031 0.0250 0.7897 0.0048 clutter 0.3340 0.2468 0.1343 0.0551 0.0093 0.2206 Precision/Correctness 0.812 0.868 0.754 0.896 0.873 0.821 Recall/Completeness 0.882 0.980 0.852 0.694 0.790 0.221 F1 0.845 0.921 0.800 0.782 0.829 0.348 Overall accuracy 0.826 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 0.9126 0.0379 0.0365 0.0109 0.0004 0.0018 building 0.0105 0.9849 0.0004 0.0040 0.0000 0.0003 low_veg 0.0420 0.0315 0.8839 0.0417 0.0000 0.0008 tree 0.0464 0.0134 0.2212 0.7156 0.0014 0.0020 car 0.0877 0.0170 0.0015 0.0186 0.8703 0.0049 clutter 0.3311 0.2381 0.1029 0.0501 0.0100 0.2679 Precision/Correctness 0.843 0.897 0.775 0.923 0.921 0.847 Recall/Completeness 0.913 0.985 0.884 0.716 0.870 0.268 F1 0.876 0.939 0.826 0.806 0.895 0.407 Overall accuracy 0.853 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 0.9026 0.0400 0.0304 0.0239 0.0021 0.0011 building 0.0177 0.9704 0.0038 0.0079 0.0001 0.0002 low_veg 0.0618 0.0655 0.7431 0.1279 0.0002 0.0015 tree 0.0680 0.0083 0.0732 0.8482 0.0018 0.0005 car 0.1588 0.0158 0.0069 0.0253 0.7925 0.0007 clutter 0.5320 0.1957 0.1351 0.0963 0.0075 0.0334 Precision/Correctness 0.813 0.899 0.787 0.754 0.925 0.770 Recall/Completeness 0.903 0.970 0.743 0.848 0.793 0.033 F1 0.855 0.933 0.765 0.798 0.853 0.064 Overall accuracy 0.834 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 0.9265 0.0307 0.0234 0.0182 0.0003 0.0009 building 0.0148 0.9761 0.0026 0.0064 0.0001 0.0001 low_veg 0.0492 0.0513 0.7809 0.1171 0.0001 0.0013 tree 0.0550 0.0057 0.0607 0.8769 0.0012 0.0005 car 0.0672 0.0149 0.0018 0.0170 0.8984 0.0007 clutter 0.5680 0.1744 0.1227 0.0908 0.0080 0.0361 Precision/Correctness 0.837 0.924 0.823 0.789 0.954 0.792 Recall/Completeness 0.926 0.976 0.781 0.877 0.898 0.036 F1 0.880 0.949 0.802 0.831 0.925 0.069 Overall accuracy 0.863 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 0.8868 0.0330 0.0586 0.0134 0.0039 0.0044 building 0.0171 0.9733 0.0017 0.0069 0.0002 0.0008 low_veg 0.1305 0.0356 0.7205 0.1080 0.0004 0.0050 tree 0.0701 0.0111 0.1627 0.7501 0.0025 0.0034 car 0.1405 0.0166 0.0044 0.0238 0.8082 0.0065 clutter 0.6532 0.0675 0.1007 0.0189 0.0019 0.1578 Precision/Correctness 0.664 0.884 0.747 0.815 0.912 0.829 Recall/Completeness 0.887 0.973 0.720 0.750 0.808 0.158 F1 0.759 0.927 0.734 0.781 0.857 0.265 Overall accuracy 0.766 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 0.9064 0.0263 0.0530 0.0097 0.0006 0.0041 building 0.0139 0.9790 0.0008 0.0054 0.0002 0.0007 low_veg 0.1163 0.0279 0.7477 0.1037 0.0001 0.0045 tree 0.0609 0.0091 0.1456 0.7791 0.0020 0.0032 car 0.0734 0.0177 0.0020 0.0179 0.8825 0.0064 clutter 0.6738 0.0624 0.0844 0.0150 0.0014 0.1631 Precision/Correctness 0.675 0.907 0.771 0.838 0.958 0.850 Recall/Completeness 0.906 0.979 0.748 0.779 0.883 0.163 F1 0.774 0.942 0.759 0.807 0.919 0.274 Overall accuracy 0.786 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 0.9154 0.0435 0.0311 0.0060 0.0023 0.0016 building 0.0118 0.9739 0.0058 0.0072 0.0002 0.0011 low_veg 0.0768 0.0664 0.8109 0.0432 0.0006 0.0020 tree 0.1312 0.0143 0.1836 0.6611 0.0062 0.0037 car 0.1687 0.0219 0.0071 0.0076 0.7937 0.0010 clutter 0.3646 0.3290 0.1278 0.0475 0.0103 0.1208 Precision/Correctness 0.868 0.889 0.769 0.868 0.902 0.669 Recall/Completeness 0.915 0.974 0.811 0.661 0.794 0.121 F1 0.891 0.930 0.790 0.750 0.845 0.205 Overall accuracy 0.858 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 0.9345 0.0357 0.0245 0.0038 0.0004 0.0011 building 0.0095 0.9782 0.0050 0.0063 0.0001 0.0010 low_veg 0.0550 0.0541 0.8532 0.0364 0.0001 0.0012 tree 0.1169 0.0117 0.1642 0.6985 0.0050 0.0036 car 0.0819 0.0241 0.0015 0.0032 0.8884 0.0008 clutter 0.3599 0.3460 0.1078 0.0462 0.0099 0.1300 Precision/Correctness 0.897 0.912 0.799 0.897 0.941 0.695 Recall/Completeness 0.934 0.978 0.853 0.699 0.888 0.130 F1 0.915 0.944 0.825 0.785 0.914 0.219 Overall accuracy 0.886 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 0.8940 0.0424 0.0370 0.0232 0.0018 0.0016 building 0.0317 0.9556 0.0058 0.0064 0.0000 0.0005 low_veg 0.0632 0.0274 0.8319 0.0747 0.0006 0.0022 tree 0.1050 0.0085 0.1893 0.6931 0.0024 0.0016 car 0.1863 0.0152 0.0012 0.0074 0.7865 0.0035 clutter 0.3069 0.3922 0.1446 0.0459 0.0099 0.1005 Precision/Correctness 0.859 0.891 0.791 0.771 0.938 0.699 Recall/Completeness 0.894 0.956 0.832 0.693 0.786 0.100 F1 0.876 0.922 0.811 0.730 0.856 0.176 Overall accuracy 0.846 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 0.9145 0.0339 0.0318 0.0186 0.0002 0.0010 building 0.0298 0.9595 0.0050 0.0053 0.0000 0.0004 low_veg 0.0579 0.0205 0.8525 0.0671 0.0005 0.0015 tree 0.0894 0.0065 0.1774 0.7232 0.0020 0.0015 car 0.0910 0.0173 0.0002 0.0051 0.8828 0.0035 clutter 0.2877 0.4071 0.1345 0.0467 0.0113 0.1127 Precision/Correctness 0.884 0.915 0.815 0.801 0.964 0.736 Recall/Completeness 0.914 0.959 0.852 0.723 0.883 0.113 F1 0.899 0.937 0.833 0.760 0.922 0.195 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9237 0.0212 0.0308 0.0195 0.0018 0.0030 building 0.0172 0.9754 0.0024 0.0040 0.0001 0.0007 low_veg 0.0933 0.0392 0.7843 0.0805 0.0004 0.0023 tree 0.0988 0.0114 0.1137 0.7699 0.0043 0.0019 car 0.1776 0.0092 0.0068 0.0185 0.7851 0.0029 clutter 0.3197 0.2321 0.1463 0.0779 0.0082 0.2158 Precision/Correctness 0.784 0.880 0.803 0.823 0.887 0.896 Recall/Completeness 0.924 0.975 0.784 0.770 0.785 0.216 F1 0.848 0.925 0.794 0.795 0.833 0.348 Overall accuracy 0.825 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 0.9426 0.0162 0.0229 0.0153 0.0003 0.0027 building 0.0155 0.9787 0.0016 0.0034 0.0001 0.0007 low_veg 0.0779 0.0312 0.8145 0.0744 0.0002 0.0019 tree 0.0864 0.0097 0.1049 0.7938 0.0036 0.0016 car 0.0828 0.0086 0.0023 0.0140 0.8889 0.0033 clutter 0.3191 0.1951 0.1474 0.0810 0.0082 0.2491 Precision/Correctness 0.812 0.909 0.825 0.845 0.915 0.913 Recall/Completeness 0.943 0.979 0.815 0.794 0.889 0.249 F1 0.872 0.943 0.820 0.819 0.902 0.391 Overall accuracy 0.851 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 0.9133 0.0189 0.0416 0.0198 0.0018 0.0046 building 0.0294 0.9464 0.0102 0.0125 0.0003 0.0011 low_veg 0.0717 0.0364 0.7824 0.1034 0.0001 0.0059 tree 0.0687 0.0099 0.1361 0.7821 0.0026 0.0005 car 0.1765 0.0106 0.0061 0.0139 0.7924 0.0003 clutter 0.4617 0.1568 0.1366 0.0247 0.0085 0.2118 Precision/Correctness 0.854 0.903 0.715 0.765 0.923 0.831 Recall/Completeness 0.913 0.946 0.782 0.782 0.792 0.212 F1 0.882 0.924 0.747 0.774 0.853 0.338 Overall accuracy 0.835 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 0.9304 0.0130 0.0358 0.0159 0.0005 0.0044 building 0.0246 0.9544 0.0091 0.0108 0.0003 0.0008 low_veg 0.0514 0.0286 0.8168 0.0973 0.0000 0.0059 tree 0.0578 0.0076 0.1261 0.8059 0.0021 0.0004 car 0.0827 0.0115 0.0017 0.0113 0.8926 0.0003 clutter 0.4839 0.1126 0.1353 0.0213 0.0087 0.2381 Precision/Correctness 0.875 0.933 0.736 0.795 0.946 0.845 Recall/Completeness 0.930 0.954 0.817 0.806 0.893 0.238 F1 0.902 0.944 0.774 0.800 0.919 0.372 Overall accuracy 0.860 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 0.9209 0.0405 0.0258 0.0105 0.0013 0.0011 building 0.0308 0.9522 0.0114 0.0044 0.0001 0.0012 low_veg 0.0256 0.0279 0.9151 0.0312 0.0000 0.0002 tree 0.1471 0.0074 0.1870 0.6565 0.0016 0.0004 car 0.1341 0.0614 0.0012 0.0031 0.7981 0.0022 clutter 0.6376 0.0488 0.1963 0.0270 0.0051 0.0852 Precision/Correctness 0.884 0.904 0.848 0.858 0.922 0.680 Recall/Completeness 0.921 0.952 0.915 0.657 0.798 0.085 F1 0.902 0.927 0.880 0.744 0.855 0.151 Overall accuracy 0.877 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 0.9381 0.0350 0.0187 0.0073 0.0002 0.0006 building 0.0281 0.9569 0.0103 0.0035 0.0001 0.0011 low_veg 0.0183 0.0208 0.9346 0.0262 0.0000 0.0001 tree 0.1260 0.0058 0.1776 0.6889 0.0013 0.0003 car 0.0485 0.0715 0.0009 0.0023 0.8747 0.0021 clutter 0.6723 0.0385 0.1790 0.0254 0.0041 0.0808 Precision/Correctness 0.902 0.922 0.868 0.885 0.963 0.720 Recall/Completeness 0.938 0.957 0.935 0.689 0.875 0.081 F1 0.920 0.939 0.900 0.775 0.917 0.145 Overall accuracy 0.897 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 0.9292 0.0262 0.0293 0.0107 0.0037 0.0008 building 0.0524 0.9385 0.0024 0.0049 0.0005 0.0013 low_veg 0.0665 0.0393 0.8284 0.0644 0.0005 0.0008 tree 0.0747 0.0079 0.1461 0.7656 0.0051 0.0005 car 0.1437 0.0081 0.0045 0.0132 0.8296 0.0009 clutter 0.4264 0.2552 0.1882 0.0504 0.0164 0.0634 Precision/Correctness 0.884 0.924 0.807 0.809 0.911 0.618 Recall/Completeness 0.929 0.939 0.828 0.766 0.830 0.063 F1 0.906 0.931 0.817 0.787 0.869 0.115 Overall accuracy 0.878 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 0.9477 0.0197 0.0233 0.0081 0.0004 0.0007 building 0.0496 0.9435 0.0014 0.0041 0.0003 0.0011 low_veg 0.0501 0.0288 0.8628 0.0577 0.0001 0.0005 tree 0.0635 0.0061 0.1332 0.7923 0.0043 0.0005 car 0.0611 0.0063 0.0016 0.0100 0.9202 0.0008 clutter 0.4223 0.2746 0.1755 0.0470 0.0154 0.0653 Precision/Correctness 0.905 0.941 0.832 0.839 0.959 0.642 Recall/Completeness 0.948 0.944 0.863 0.792 0.920 0.065 F1 0.926 0.942 0.847 0.815 0.939 0.118 Overall accuracy 0.900 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 0.9091 0.0183 0.0509 0.0135 0.0046 0.0035 building 0.0192 0.9638 0.0054 0.0104 0.0003 0.0009 low_veg 0.0643 0.0206 0.8513 0.0556 0.0008 0.0074 tree 0.0438 0.0155 0.2486 0.6859 0.0044 0.0019 car 0.1113 0.0147 0.0125 0.0136 0.8422 0.0057 clutter 0.3449 0.0316 0.2352 0.0324 0.0111 0.3449 Precision/Correctness 0.884 0.938 0.632 0.855 0.887 0.869 Recall/Completeness 0.909 0.964 0.851 0.686 0.842 0.345 F1 0.897 0.951 0.726 0.761 0.864 0.494 Overall accuracy 0.843 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 0.9322 0.0135 0.0408 0.0100 0.0007 0.0028 building 0.0150 0.9710 0.0039 0.0091 0.0002 0.0008 low_veg 0.0507 0.0149 0.8776 0.0498 0.0001 0.0070 tree 0.0358 0.0141 0.2317 0.7132 0.0035 0.0017 car 0.0403 0.0160 0.0043 0.0096 0.9234 0.0064 clutter 0.3433 0.0290 0.2240 0.0306 0.0109 0.3623 Precision/Correctness 0.904 0.953 0.651 0.881 0.938 0.891 Recall/Completeness 0.932 0.971 0.878 0.713 0.923 0.362 F1 0.918 0.962 0.747 0.788 0.930 0.515 Overall accuracy 0.866 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 0.8888 0.0714 0.0142 0.0209 0.0015 0.0033 building 0.0249 0.9606 0.0027 0.0045 0.0005 0.0068 low_veg 0.0506 0.0468 0.7450 0.1561 0.0004 0.0011 tree 0.0320 0.0196 0.1597 0.7879 0.0006 0.0002 car 0.1397 0.0349 0.0043 0.0272 0.7906 0.0032 clutter 0.4099 0.0334 0.0898 0.0336 0.0054 0.4279 Precision/Correctness 0.894 0.826 0.727 0.667 0.885 0.921 Recall/Completeness 0.889 0.961 0.745 0.788 0.791 0.428 F1 0.892 0.888 0.736 0.723 0.835 0.584 Overall accuracy 0.840 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 0.9032 0.0679 0.0102 0.0154 0.0002 0.0031 building 0.0233 0.9641 0.0022 0.0034 0.0004 0.0066 low_veg 0.0389 0.0416 0.7688 0.1498 0.0001 0.0008 tree 0.0181 0.0183 0.1513 0.8119 0.0003 0.0001 car 0.0655 0.0402 0.0031 0.0199 0.8684 0.0030 clutter 0.4034 0.0311 0.0893 0.0335 0.0049 0.4377 Precision/Correctness 0.904 0.838 0.748 0.694 0.930 0.927 Recall/Completeness 0.903 0.964 0.769 0.812 0.868 0.438 F1 0.903 0.897 0.758 0.749 0.898 0.595 Overall accuracy 0.854 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image