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 0.930 0.013 0.031 0.016 0.002 0.008 building 0.015 0.971 0.005 0.004 0.000 0.005 low_veg 0.040 0.009 0.883 0.055 0.000 0.013 tree 0.035 0.006 0.119 0.835 0.002 0.003 car 0.066 0.003 0.002 0.013 0.909 0.006 clutter 0.298 0.113 0.138 0.021 0.010 0.420 Precision/Correctness 0.899 0.950 0.831 0.885 0.919 0.731 Recall/Completeness 0.930 0.971 0.883 0.835 0.909 0.420 F1 0.914 0.961 0.856 0.859 0.914 0.534 Overall accuracy 0.890	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 0.947 0.009 0.025 0.011 0.000 0.007 building 0.012 0.978 0.003 0.003 0.000 0.004 low_veg 0.030 0.004 0.907 0.047 0.000 0.012 tree 0.026 0.004 0.104 0.862 0.002 0.003 car 0.009 0.004 0.001 0.009 0.971 0.007 clutter 0.298 0.099 0.125 0.017 0.010 0.452 Precision/Correctness 0.919 0.965 0.855 0.910 0.949 0.750 Recall/Completeness 0.947 0.978 0.907 0.862 0.971 0.452 F1 0.933 0.972 0.880 0.885 0.960 0.564 Overall accuracy 0.911
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9084 0.0133 0.0417 0.0113 0.0022 0.0230 building 0.0072 0.9591 0.0088 0.0201 0.0000 0.0048 low_veg 0.0260 0.0067 0.8893 0.0375 0.0002 0.0403 tree 0.0378 0.0093 0.1260 0.8214 0.0009 0.0046 car 0.0679 0.0004 0.0009 0.0074 0.9190 0.0044 clutter 0.1243 0.1738 0.1310 0.0206 0.0137 0.5367 Precision/Correctness 0.913 0.942 0.861 0.912 0.909 0.342 Recall/Completeness 0.908 0.959 0.889 0.821 0.919 0.537 F1 0.911 0.951 0.875 0.864 0.914 0.418 Overall accuracy 0.885 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 0.9296 0.0082 0.0333 0.0073 0.0001 0.0214 building 0.0040 0.9669 0.0055 0.0197 0.0000 0.0039 low_veg 0.0187 0.0032 0.9086 0.0279 0.0001 0.0415 tree 0.0291 0.0076 0.1054 0.8537 0.0005 0.0036 car 0.0197 0.0002 0.0000 0.0024 0.9737 0.0040 clutter 0.0861 0.1721 0.0904 0.0137 0.0169 0.6208 Precision/Correctness 0.940 0.963 0.887 0.933 0.954 0.313 Recall/Completeness 0.930 0.967 0.909 0.854 0.974 0.621 F1 0.935 0.965 0.898 0.892 0.964 0.416 Overall accuracy 0.909 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 0.9083 0.0064 0.0699 0.0109 0.0013 0.0033 building 0.0056 0.9828 0.0054 0.0049 0.0000 0.0012 low_veg 0.0105 0.0032 0.9540 0.0300 0.0000 0.0022 tree 0.0163 0.0040 0.1742 0.8043 0.0007 0.0006 car 0.0486 0.0001 0.0016 0.0081 0.9416 0.0000 clutter 0.0604 0.2990 0.1813 0.0541 0.0019 0.4033 Precision/Correctness 0.927 0.934 0.842 0.951 0.926 0.457 Recall/Completeness 0.908 0.983 0.954 0.804 0.942 0.403 F1 0.917 0.958 0.895 0.872 0.934 0.429 Overall accuracy 0.894 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 0.9349 0.0034 0.0532 0.0063 0.0000 0.0022 building 0.0035 0.9913 0.0015 0.0029 0.0000 0.0009 low_veg 0.0059 0.0015 0.9677 0.0232 0.0000 0.0017 tree 0.0123 0.0032 0.1570 0.8267 0.0004 0.0004 car 0.0012 0.0000 0.0000 0.0035 0.9953 0.0000 clutter 0.0303 0.3380 0.1543 0.0607 0.0008 0.4159 Precision/Correctness 0.951 0.957 0.861 0.965 0.967 0.464 Recall/Completeness 0.935 0.991 0.968 0.827 0.995 0.416 F1 0.943 0.974 0.911 0.890 0.981 0.439 Overall accuracy 0.912 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 0.9136 0.0195 0.0337 0.0170 0.0029 0.0132 building 0.0095 0.9827 0.0033 0.0028 0.0000 0.0016 low_veg 0.0506 0.0177 0.8644 0.0565 0.0001 0.0108 tree 0.0291 0.0066 0.0908 0.8683 0.0020 0.0031 car 0.0582 0.0062 0.0006 0.0183 0.9155 0.0012 clutter 0.2371 0.1447 0.2126 0.0244 0.0053 0.3759 Precision/Correctness 0.893 0.945 0.850 0.914 0.921 0.562 Recall/Completeness 0.914 0.983 0.864 0.868 0.915 0.376 F1 0.903 0.964 0.857 0.890 0.918 0.450 Overall accuracy 0.896 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 0.9398 0.0128 0.0250 0.0104 0.0001 0.0119 building 0.0074 0.9883 0.0012 0.0018 0.0000 0.0012 low_veg 0.0396 0.0079 0.8989 0.0453 0.0000 0.0083 tree 0.0210 0.0050 0.0748 0.8954 0.0015 0.0025 car 0.0054 0.0069 0.0000 0.0123 0.9745 0.0009 clutter 0.2315 0.1379 0.1891 0.0181 0.0059 0.4175 Precision/Correctness 0.921 0.967 0.880 0.939 0.965 0.572 Recall/Completeness 0.940 0.988 0.899 0.895 0.974 0.418 F1 0.930 0.977 0.889 0.917 0.970 0.483 Overall accuracy 0.924 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 0.8999 0.0147 0.0558 0.0201 0.0029 0.0067 building 0.0083 0.9837 0.0033 0.0029 0.0001 0.0017 low_veg 0.0270 0.0091 0.9169 0.0413 0.0000 0.0056 tree 0.0227 0.0046 0.1901 0.7770 0.0014 0.0042 car 0.0558 0.0235 0.0015 0.0263 0.8868 0.0061 clutter 0.2065 0.1653 0.1610 0.0264 0.0030 0.4378 Precision/Correctness 0.896 0.947 0.789 0.921 0.906 0.749 Recall/Completeness 0.900 0.984 0.917 0.777 0.887 0.438 F1 0.898 0.965 0.848 0.843 0.896 0.552 Overall accuracy 0.877 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 0.9289 0.0079 0.0442 0.0136 0.0004 0.0050 building 0.0066 0.9894 0.0008 0.0021 0.0001 0.0011 low_veg 0.0174 0.0039 0.9436 0.0310 0.0000 0.0041 tree 0.0174 0.0035 0.1743 0.8001 0.0012 0.0035 car 0.0065 0.0268 0.0002 0.0175 0.9425 0.0064 clutter 0.2000 0.1553 0.1341 0.0197 0.0035 0.4874 Precision/Correctness 0.923 0.967 0.811 0.945 0.947 0.787 Recall/Completeness 0.929 0.989 0.944 0.800 0.943 0.487 F1 0.926 0.978 0.872 0.867 0.945 0.602 Overall accuracy 0.902 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 0.9401 0.0067 0.0277 0.0218 0.0014 0.0023 building 0.0086 0.9787 0.0059 0.0039 0.0000 0.0029 low_veg 0.0422 0.0074 0.8410 0.1039 0.0001 0.0053 tree 0.0275 0.0025 0.0405 0.9268 0.0015 0.0012 car 0.0611 0.0005 0.0010 0.0203 0.9154 0.0017 clutter 0.4911 0.1044 0.1762 0.0575 0.0128 0.1579 Precision/Correctness 0.858 0.969 0.816 0.813 0.930 0.803 Recall/Completeness 0.940 0.979 0.841 0.927 0.915 0.158 F1 0.897 0.974 0.828 0.866 0.923 0.264 Overall accuracy 0.883 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 0.9575 0.0034 0.0218 0.0154 0.0002 0.0016 building 0.0045 0.9880 0.0024 0.0026 0.0000 0.0024 low_veg 0.0320 0.0027 0.8675 0.0933 0.0000 0.0044 tree 0.0162 0.0013 0.0287 0.9522 0.0009 0.0008 car 0.0041 0.0001 0.0001 0.0123 0.9818 0.0016 clutter 0.5134 0.0917 0.1629 0.0495 0.0141 0.1684 Precision/Correctness 0.881 0.980 0.851 0.847 0.945 0.831 Recall/Completeness 0.958 0.988 0.868 0.952 0.982 0.168 F1 0.918 0.984 0.859 0.897 0.963 0.280 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 0.9007 0.0242 0.0455 0.0137 0.0029 0.0130 building 0.0076 0.9799 0.0032 0.0050 0.0000 0.0043 low_veg 0.0837 0.0142 0.8214 0.0532 0.0002 0.0273 tree 0.0314 0.0048 0.1231 0.8291 0.0024 0.0093 car 0.0572 0.0047 0.0052 0.0193 0.9070 0.0067 clutter 0.4329 0.0424 0.1158 0.0091 0.0003 0.3996 Precision/Correctness 0.764 0.931 0.801 0.896 0.943 0.762 Recall/Completeness 0.901 0.980 0.821 0.829 0.907 0.400 F1 0.827 0.955 0.811 0.861 0.924 0.524 Overall accuracy 0.835 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 0.9175 0.0201 0.0412 0.0092 0.0003 0.0118 building 0.0048 0.9861 0.0014 0.0039 0.0000 0.0037 low_veg 0.0723 0.0086 0.8479 0.0448 0.0000 0.0263 tree 0.0230 0.0033 0.1046 0.8590 0.0018 0.0082 car 0.0107 0.0044 0.0058 0.0141 0.9585 0.0065 clutter 0.4440 0.0384 0.1035 0.0066 0.0002 0.4074 Precision/Correctness 0.777 0.947 0.822 0.920 0.976 0.783 Recall/Completeness 0.917 0.986 0.848 0.859 0.958 0.407 F1 0.841 0.966 0.835 0.889 0.967 0.536 Overall accuracy 0.855 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 0.9230 0.0158 0.0504 0.0048 0.0021 0.0038 building 0.0038 0.9864 0.0064 0.0018 0.0000 0.0016 low_veg 0.0583 0.0180 0.8908 0.0272 0.0004 0.0053 tree 0.0707 0.0074 0.1367 0.7728 0.0068 0.0056 car 0.0723 0.0008 0.0029 0.0030 0.9158 0.0051 clutter 0.3504 0.2299 0.1241 0.0099 0.0052 0.2805 Precision/Correctness 0.909 0.950 0.783 0.934 0.923 0.698 Recall/Completeness 0.923 0.986 0.891 0.773 0.916 0.280 F1 0.916 0.968 0.834 0.846 0.919 0.400 Overall accuracy 0.899 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 0.9378 0.0099 0.0464 0.0029 0.0002 0.0028 building 0.0017 0.9914 0.0046 0.0012 0.0000 0.0011 low_veg 0.0380 0.0090 0.9279 0.0216 0.0000 0.0035 tree 0.0533 0.0056 0.1150 0.8161 0.0050 0.0050 car 0.0112 0.0003 0.0003 0.0015 0.9807 0.0060 clutter 0.3620 0.2350 0.1106 0.0073 0.0048 0.2803 Precision/Correctness 0.935 0.967 0.807 0.954 0.957 0.721 Recall/Completeness 0.938 0.991 0.928 0.816 0.981 0.280 F1 0.937 0.979 0.863 0.880 0.969 0.404 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 0.9387 0.0114 0.0244 0.0188 0.0014 0.0053 building 0.0075 0.9845 0.0039 0.0028 0.0000 0.0012 low_veg 0.0662 0.0083 0.8496 0.0694 0.0000 0.0065 tree 0.0513 0.0043 0.1052 0.8351 0.0026 0.0016 car 0.0881 0.0004 0.0002 0.0091 0.8912 0.0110 clutter 0.2297 0.2485 0.1475 0.0153 0.0465 0.3126 Precision/Correctness 0.910 0.953 0.857 0.831 0.909 0.724 Recall/Completeness 0.939 0.985 0.850 0.835 0.891 0.313 F1 0.924 0.968 0.853 0.833 0.900 0.437 Overall accuracy 0.900 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 0.9560 0.0062 0.0201 0.0137 0.0001 0.0039 building 0.0039 0.9910 0.0021 0.0021 0.0000 0.0008 low_veg 0.0589 0.0039 0.8713 0.0607 0.0000 0.0052 tree 0.0353 0.0028 0.0910 0.8676 0.0020 0.0013 car 0.0109 0.0002 0.0000 0.0070 0.9694 0.0126 clutter 0.2025 0.2402 0.1385 0.0133 0.0583 0.3472 Precision/Correctness 0.935 0.970 0.884 0.859 0.923 0.745 Recall/Completeness 0.956 0.991 0.871 0.868 0.969 0.347 F1 0.946 0.980 0.878 0.863 0.946 0.474 Overall accuracy 0.924 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 0.9419 0.0047 0.0204 0.0238 0.0010 0.0083 building 0.0171 0.9764 0.0030 0.0018 0.0001 0.0017 low_veg 0.0499 0.0076 0.8407 0.0868 0.0003 0.0148 tree 0.0432 0.0074 0.0680 0.8723 0.0037 0.0054 car 0.0808 0.0081 0.0098 0.0116 0.8731 0.0165 clutter 0.2224 0.1951 0.1369 0.0384 0.0058 0.4014 Precision/Correctness 0.863 0.929 0.859 0.846 0.924 0.810 Recall/Completeness 0.942 0.976 0.841 0.872 0.873 0.401 F1 0.901 0.952 0.850 0.859 0.898 0.537 Overall accuracy 0.876 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 0.9590 0.0021 0.0138 0.0179 0.0001 0.0072 building 0.0151 0.9809 0.0010 0.0014 0.0000 0.0016 low_veg 0.0356 0.0044 0.8646 0.0814 0.0001 0.0140 tree 0.0326 0.0062 0.0588 0.8950 0.0028 0.0045 car 0.0073 0.0093 0.0101 0.0071 0.9482 0.0179 clutter 0.2111 0.1694 0.1244 0.0359 0.0059 0.4532 Precision/Correctness 0.890 0.948 0.885 0.868 0.942 0.831 Recall/Completeness 0.959 0.981 0.865 0.895 0.948 0.453 F1 0.923 0.964 0.875 0.881 0.945 0.586 Overall accuracy 0.899 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 0.9276 0.0084 0.0418 0.0186 0.0011 0.0025 building 0.0120 0.9725 0.0076 0.0041 0.0003 0.0035 low_veg 0.0396 0.0048 0.8301 0.1071 0.0001 0.0184 tree 0.0315 0.0041 0.0646 0.8959 0.0024 0.0014 car 0.1041 0.0020 0.0020 0.0130 0.8711 0.0079 clutter 0.3399 0.1271 0.1103 0.0160 0.0100 0.3967 Precision/Correctness 0.905 0.947 0.778 0.805 0.939 0.851 Recall/Completeness 0.928 0.973 0.830 0.896 0.871 0.397 F1 0.916 0.960 0.803 0.848 0.904 0.541 Overall accuracy 0.882 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 0.9415 0.0048 0.0374 0.0139 0.0004 0.0019 building 0.0078 0.9798 0.0060 0.0030 0.0002 0.0032 low_veg 0.0245 0.0009 0.8567 0.0995 0.0000 0.0183 tree 0.0216 0.0021 0.0548 0.9185 0.0018 0.0012 car 0.0201 0.0019 0.0001 0.0108 0.9577 0.0094 clutter 0.3435 0.0948 0.1011 0.0134 0.0109 0.4363 Precision/Correctness 0.925 0.967 0.799 0.835 0.947 0.865 Recall/Completeness 0.942 0.980 0.857 0.919 0.958 0.436 F1 0.933 0.973 0.827 0.875 0.952 0.580 Overall accuracy 0.904 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 0.9380 0.0233 0.0166 0.0133 0.0013 0.0075 building 0.0295 0.9561 0.0104 0.0011 0.0000 0.0028 low_veg 0.0175 0.0065 0.9469 0.0277 0.0000 0.0014 tree 0.0667 0.0029 0.0740 0.8542 0.0011 0.0010 car 0.0523 0.0008 0.0003 0.0025 0.9436 0.0005 clutter 0.4450 0.0446 0.1688 0.0393 0.0009 0.3014 Precision/Correctness 0.924 0.952 0.914 0.889 0.948 0.608 Recall/Completeness 0.938 0.956 0.947 0.854 0.944 0.301 F1 0.931 0.954 0.930 0.871 0.946 0.403 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_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9520 0.0206 0.0129 0.0083 0.0001 0.0061 building 0.0259 0.9643 0.0069 0.0006 0.0000 0.0024 low_veg 0.0097 0.0041 0.9631 0.0221 0.0000 0.0010 tree 0.0476 0.0016 0.0620 0.8873 0.0007 0.0008 car 0.0033 0.0010 0.0001 0.0014 0.9937 0.0006 clutter 0.4712 0.0397 0.1572 0.0381 0.0004 0.2933 Precision/Correctness 0.940 0.961 0.932 0.918 0.987 0.625 Recall/Completeness 0.952 0.964 0.963 0.887 0.994 0.293 F1 0.946 0.963 0.947 0.902 0.990 0.399 Overall accuracy 0.939 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 0.9592 0.0058 0.0205 0.0094 0.0035 0.0016 building 0.0468 0.9448 0.0040 0.0009 0.0008 0.0027 low_veg 0.0489 0.0070 0.8859 0.0537 0.0004 0.0041 tree 0.0468 0.0098 0.0757 0.8629 0.0043 0.0005 car 0.0489 0.0005 0.0007 0.0095 0.9357 0.0047 clutter 0.2803 0.2999 0.2340 0.0193 0.0112 0.1553 Precision/Correctness 0.917 0.962 0.862 0.863 0.926 0.613 Recall/Completeness 0.959 0.945 0.886 0.863 0.936 0.155 F1 0.937 0.953 0.874 0.863 0.931 0.248 Overall accuracy 0.915 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 0.9725 0.0031 0.0162 0.0067 0.0002 0.0012 building 0.0425 0.9524 0.0015 0.0005 0.0007 0.0025 low_veg 0.0329 0.0035 0.9133 0.0466 0.0000 0.0037 tree 0.0348 0.0084 0.0628 0.8904 0.0032 0.0004 car 0.0036 0.0002 0.0001 0.0061 0.9849 0.0051 clutter 0.2584 0.3364 0.2319 0.0158 0.0104 0.1472 Precision/Correctness 0.934 0.969 0.885 0.892 0.967 0.605 Recall/Completeness 0.973 0.952 0.913 0.890 0.985 0.147 F1 0.953 0.961 0.899 0.891 0.976 0.237 Overall accuracy 0.933 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 0.9387 0.0050 0.0282 0.0091 0.0038 0.0152 building 0.0093 0.9822 0.0034 0.0023 0.0000 0.0028 low_veg 0.0432 0.0025 0.8773 0.0435 0.0009 0.0326 tree 0.0372 0.0048 0.1336 0.8169 0.0035 0.0040 car 0.0550 0.0018 0.0037 0.0098 0.9222 0.0075 clutter 0.1520 0.0166 0.1693 0.0056 0.0129 0.6436 Precision/Correctness 0.934 0.983 0.754 0.920 0.907 0.761 Recall/Completeness 0.939 0.982 0.877 0.817 0.922 0.644 F1 0.936 0.982 0.811 0.865 0.914 0.697 Overall accuracy 0.903 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 0.9554 0.0029 0.0219 0.0060 0.0001 0.0137 building 0.0049 0.9904 0.0009 0.0016 0.0000 0.0022 low_veg 0.0286 0.0003 0.9045 0.0345 0.0001 0.0321 tree 0.0272 0.0035 0.1168 0.8467 0.0026 0.0033 car 0.0065 0.0020 0.0003 0.0065 0.9769 0.0077 clutter 0.1459 0.0161 0.1616 0.0037 0.0134 0.6594 Precision/Correctness 0.952 0.989 0.777 0.944 0.950 0.782 Recall/Completeness 0.955 0.990 0.904 0.847 0.977 0.659 F1 0.954 0.990 0.836 0.893 0.963 0.716 Overall accuracy 0.923 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 0.9349 0.0223 0.0077 0.0230 0.0014 0.0106 building 0.0263 0.9308 0.0023 0.0014 0.0017 0.0375 low_veg 0.0533 0.0140 0.7951 0.1347 0.0002 0.0026 tree 0.0311 0.0091 0.0908 0.8679 0.0009 0.0001 car 0.0454 0.0085 0.0003 0.0242 0.9198 0.0019 clutter 0.1996 0.0194 0.0776 0.0020 0.0124 0.6889 Precision/Correctness 0.935 0.933 0.814 0.723 0.847 0.816 Recall/Completeness 0.935 0.931 0.795 0.868 0.920 0.689 F1 0.935 0.932 0.804 0.789 0.882 0.747 Overall accuracy 0.892 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 0.9490 0.0188 0.0056 0.0164 0.0002 0.0100 building 0.0236 0.9353 0.0015 0.0008 0.0016 0.0371 low_veg 0.0412 0.0099 0.8198 0.1270 0.0000 0.0021 tree 0.0224 0.0071 0.0825 0.8875 0.0005 0.0001 car 0.0033 0.0087 0.0000 0.0179 0.9689 0.0011 clutter 0.1921 0.0175 0.0762 0.0016 0.0116 0.7010 Precision/Correctness 0.943 0.944 0.830 0.757 0.870 0.824 Recall/Completeness 0.949 0.935 0.820 0.887 0.969 0.701 F1 0.946 0.940 0.825 0.817 0.917 0.758 Overall accuracy 0.906 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image