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 0.906 0.024 0.035 0.017 0.008 0.010 building 0.029 0.949 0.013 0.003 0.000 0.005 low_veg 0.060 0.013 0.839 0.075 0.000 0.013 tree 0.041 0.005 0.115 0.833 0.002 0.003 car 0.161 0.008 0.005 0.016 0.799 0.009 clutter 0.262 0.126 0.182 0.022 0.005 0.402 Precision/Correctness 0.872 0.930 0.806 0.859 0.831 0.706 Recall/Completeness 0.906 0.949 0.839 0.833 0.799 0.402 F1 0.889 0.940 0.822 0.846 0.815 0.512 Overall accuracy 0.865	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 0.928 0.019 0.028 0.012 0.003 0.009 building 0.025 0.958 0.010 0.002 0.000 0.004 low_veg 0.045 0.007 0.871 0.065 0.000 0.012 tree 0.030 0.004 0.101 0.861 0.002 0.002 car 0.087 0.008 0.003 0.011 0.881 0.010 clutter 0.252 0.110 0.171 0.018 0.005 0.443 Precision/Correctness 0.900 0.948 0.832 0.886 0.894 0.733 Recall/Completeness 0.928 0.958 0.871 0.861 0.881 0.443 F1 0.914 0.953 0.851 0.873 0.887 0.552 Overall accuracy 0.891
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8791 0.0347 0.0454 0.0167 0.0051 0.0190 building 0.0243 0.9355 0.0308 0.0050 0.0000 0.0043 low_veg 0.0436 0.0127 0.8815 0.0490 0.0003 0.0129 tree 0.0355 0.0094 0.1588 0.7923 0.0012 0.0028 car 0.1395 0.0222 0.0034 0.0128 0.7801 0.0420 clutter 0.1893 0.2114 0.1828 0.0229 0.0021 0.3915 Precision/Correctness 0.874 0.901 0.823 0.900 0.845 0.417 Recall/Completeness 0.879 0.936 0.881 0.792 0.780 0.391 F1 0.876 0.918 0.851 0.843 0.811 0.404 Overall accuracy 0.859 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 0.9060 0.0292 0.0351 0.0110 0.0017 0.0170 building 0.0191 0.9476 0.0260 0.0038 0.0000 0.0036 low_veg 0.0334 0.0078 0.9091 0.0375 0.0002 0.0120 tree 0.0266 0.0077 0.1376 0.8252 0.0008 0.0022 car 0.0681 0.0247 0.0024 0.0075 0.8489 0.0485 clutter 0.1562 0.2103 0.1498 0.0138 0.0027 0.4672 Precision/Correctness 0.907 0.926 0.852 0.928 0.919 0.414 Recall/Completeness 0.906 0.948 0.909 0.825 0.849 0.467 F1 0.906 0.937 0.880 0.873 0.883 0.439 Overall accuracy 0.889 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 0.8876 0.0108 0.0741 0.0197 0.0054 0.0025 building 0.0149 0.9573 0.0147 0.0081 0.0000 0.0049 low_veg 0.0214 0.0046 0.8993 0.0619 0.0001 0.0127 tree 0.0157 0.0026 0.1439 0.8346 0.0007 0.0026 car 0.1278 0.0151 0.0072 0.0283 0.8214 0.0001 clutter 0.1053 0.2687 0.2818 0.0629 0.0147 0.2668 Precision/Correctness 0.892 0.922 0.852 0.909 0.812 0.122 Recall/Completeness 0.888 0.957 0.899 0.835 0.821 0.267 F1 0.890 0.939 0.875 0.870 0.817 0.168 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_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9169 0.0071 0.0587 0.0133 0.0023 0.0018 building 0.0113 0.9713 0.0076 0.0056 0.0000 0.0042 low_veg 0.0130 0.0025 0.9198 0.0519 0.0000 0.0128 tree 0.0117 0.0020 0.1301 0.8534 0.0004 0.0023 car 0.0622 0.0148 0.0039 0.0210 0.8980 0.0001 clutter 0.0602 0.3022 0.2595 0.0633 0.0133 0.3015 Precision/Correctness 0.926 0.949 0.870 0.927 0.888 0.103 Recall/Completeness 0.917 0.971 0.920 0.853 0.898 0.301 F1 0.922 0.960 0.894 0.889 0.893 0.153 Overall accuracy 0.897 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 0.8725 0.0229 0.0630 0.0229 0.0093 0.0095 building 0.0192 0.9617 0.0094 0.0046 0.0000 0.0051 low_veg 0.0638 0.0245 0.8322 0.0701 0.0002 0.0091 tree 0.0300 0.0058 0.0985 0.8613 0.0018 0.0025 car 0.1509 0.0042 0.0033 0.0238 0.8131 0.0047 clutter 0.2739 0.1750 0.3066 0.0240 0.0046 0.2158 Precision/Correctness 0.861 0.933 0.799 0.892 0.827 0.444 Recall/Completeness 0.873 0.962 0.832 0.861 0.813 0.216 F1 0.867 0.947 0.815 0.876 0.820 0.290 Overall accuracy 0.866 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 0.9023 0.0158 0.0538 0.0156 0.0040 0.0085 building 0.0148 0.9711 0.0059 0.0033 0.0000 0.0048 low_veg 0.0491 0.0143 0.8709 0.0584 0.0001 0.0071 tree 0.0214 0.0043 0.0828 0.8882 0.0013 0.0021 car 0.0750 0.0040 0.0016 0.0154 0.8987 0.0054 clutter 0.2695 0.1754 0.2839 0.0163 0.0053 0.2496 Precision/Correctness 0.897 0.956 0.831 0.919 0.897 0.454 Recall/Completeness 0.902 0.971 0.871 0.888 0.899 0.250 F1 0.899 0.964 0.851 0.903 0.898 0.322 Overall accuracy 0.898 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 0.8704 0.0257 0.0624 0.0270 0.0068 0.0077 building 0.0137 0.9730 0.0089 0.0029 0.0000 0.0015 low_veg 0.0377 0.0138 0.8679 0.0694 0.0001 0.0110 tree 0.0210 0.0048 0.1365 0.8340 0.0009 0.0028 car 0.1393 0.0158 0.0073 0.0395 0.7872 0.0110 clutter 0.2188 0.2092 0.2475 0.0370 0.0008 0.2867 Precision/Correctness 0.872 0.927 0.803 0.890 0.846 0.599 Recall/Completeness 0.870 0.973 0.868 0.834 0.787 0.287 F1 0.871 0.949 0.834 0.861 0.816 0.388 Overall accuracy 0.864 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 0.9027 0.0189 0.0500 0.0193 0.0026 0.0064 building 0.0110 0.9819 0.0043 0.0017 0.0000 0.0011 low_veg 0.0244 0.0079 0.9001 0.0577 0.0000 0.0099 tree 0.0155 0.0038 0.1195 0.8582 0.0008 0.0024 car 0.0683 0.0173 0.0059 0.0278 0.8690 0.0117 clutter 0.2065 0.2048 0.2208 0.0322 0.0007 0.3350 Precision/Correctness 0.907 0.948 0.832 0.915 0.915 0.633 Recall/Completeness 0.903 0.982 0.900 0.858 0.869 0.335 F1 0.905 0.965 0.865 0.886 0.892 0.438 Overall accuracy 0.894 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 0.9253 0.0111 0.0298 0.0221 0.0081 0.0035 building 0.0210 0.9634 0.0084 0.0049 0.0002 0.0021 low_veg 0.0801 0.0139 0.7965 0.1037 0.0003 0.0055 tree 0.0383 0.0044 0.0523 0.9017 0.0019 0.0014 car 0.1697 0.0019 0.0102 0.0254 0.7887 0.0042 clutter 0.5035 0.1196 0.1822 0.0623 0.0069 0.1256 Precision/Correctness 0.821 0.958 0.788 0.804 0.839 0.751 Recall/Completeness 0.925 0.963 0.796 0.902 0.789 0.126 F1 0.870 0.961 0.792 0.850 0.813 0.215 Overall accuracy 0.859 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 0.9485 0.0064 0.0231 0.0159 0.0036 0.0026 building 0.0152 0.9749 0.0044 0.0036 0.0001 0.0018 low_veg 0.0609 0.0058 0.8363 0.0930 0.0001 0.0040 tree 0.0243 0.0028 0.0388 0.9320 0.0012 0.0010 car 0.0951 0.0009 0.0057 0.0164 0.8778 0.0040 clutter 0.5205 0.1015 0.1752 0.0548 0.0077 0.1402 Precision/Correctness 0.852 0.973 0.827 0.840 0.899 0.797 Recall/Completeness 0.949 0.975 0.836 0.932 0.878 0.140 F1 0.897 0.974 0.832 0.883 0.888 0.238 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8755 0.0223 0.0707 0.0132 0.0107 0.0076 building 0.0245 0.9587 0.0063 0.0062 0.0004 0.0039 low_veg 0.1028 0.0188 0.7899 0.0718 0.0004 0.0162 tree 0.0386 0.0073 0.1298 0.8170 0.0016 0.0057 car 0.1451 0.0072 0.0029 0.0235 0.8025 0.0187 clutter 0.3789 0.0555 0.1393 0.0135 0.0010 0.4118 Precision/Correctness 0.744 0.919 0.759 0.870 0.846 0.837 Recall/Completeness 0.876 0.959 0.790 0.817 0.803 0.412 F1 0.805 0.938 0.774 0.843 0.824 0.552 Overall accuracy 0.814 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 0.8983 0.0174 0.0645 0.0081 0.0050 0.0066 building 0.0206 0.9670 0.0036 0.0049 0.0004 0.0035 low_veg 0.0857 0.0127 0.8232 0.0634 0.0001 0.0149 tree 0.0287 0.0059 0.1126 0.8467 0.0012 0.0050 car 0.0752 0.0072 0.0011 0.0173 0.8781 0.0211 clutter 0.3844 0.0507 0.1265 0.0104 0.0007 0.4273 Precision/Correctness 0.766 0.938 0.783 0.896 0.908 0.858 Recall/Completeness 0.898 0.967 0.823 0.847 0.878 0.427 F1 0.827 0.952 0.803 0.871 0.893 0.571 Overall accuracy 0.838 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 0.9028 0.0293 0.0446 0.0059 0.0078 0.0096 building 0.0510 0.9198 0.0149 0.0018 0.0003 0.0122 low_veg 0.0747 0.0224 0.8470 0.0374 0.0014 0.0171 tree 0.0774 0.0092 0.1402 0.7622 0.0069 0.0042 car 0.1679 0.0187 0.0077 0.0069 0.7902 0.0086 clutter 0.3365 0.2606 0.1856 0.0143 0.0026 0.2004 Precision/Correctness 0.856 0.926 0.762 0.913 0.826 0.353 Recall/Completeness 0.903 0.920 0.847 0.762 0.790 0.200 F1 0.879 0.923 0.802 0.831 0.808 0.256 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 0.9237 0.0235 0.0383 0.0037 0.0032 0.0076 building 0.0483 0.9268 0.0119 0.0012 0.0003 0.0115 low_veg 0.0491 0.0130 0.8927 0.0302 0.0005 0.0146 tree 0.0579 0.0070 0.1212 0.8056 0.0052 0.0032 car 0.0888 0.0224 0.0033 0.0039 0.8724 0.0092 clutter 0.3388 0.2709 0.1691 0.0113 0.0026 0.2072 Precision/Correctness 0.886 0.944 0.793 0.938 0.893 0.350 Recall/Completeness 0.924 0.927 0.893 0.806 0.872 0.207 F1 0.905 0.935 0.840 0.867 0.883 0.260 Overall accuracy 0.887 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 0.9241 0.0198 0.0238 0.0211 0.0058 0.0053 building 0.0189 0.9699 0.0048 0.0030 0.0000 0.0034 low_veg 0.1161 0.0161 0.7408 0.1209 0.0001 0.0061 tree 0.0642 0.0057 0.0795 0.8468 0.0025 0.0013 car 0.1757 0.0010 0.0008 0.0064 0.8083 0.0078 clutter 0.2987 0.2561 0.1691 0.0139 0.0233 0.2389 Precision/Correctness 0.861 0.937 0.852 0.768 0.873 0.644 Recall/Completeness 0.924 0.970 0.741 0.847 0.808 0.239 F1 0.891 0.953 0.793 0.806 0.839 0.348 Overall accuracy 0.867 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_5_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9441 0.0134 0.0204 0.0159 0.0023 0.0037 building 0.0147 0.9766 0.0035 0.0022 0.0000 0.0030 low_veg 0.1078 0.0103 0.7655 0.1116 0.0000 0.0047 tree 0.0462 0.0038 0.0700 0.8771 0.0020 0.0010 car 0.0858 0.0009 0.0003 0.0037 0.9013 0.0082 clutter 0.2691 0.2433 0.1666 0.0112 0.0287 0.2811 Precision/Correctness 0.889 0.957 0.877 0.794 0.917 0.678 Recall/Completeness 0.944 0.977 0.766 0.877 0.901 0.281 F1 0.916 0.967 0.818 0.833 0.909 0.398 Overall accuracy 0.893 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 0.9155 0.0135 0.0224 0.0286 0.0084 0.0116 building 0.0284 0.9631 0.0056 0.0019 0.0001 0.0010 low_veg 0.0731 0.0094 0.7972 0.1036 0.0004 0.0163 tree 0.0506 0.0039 0.0729 0.8661 0.0033 0.0031 car 0.1737 0.0027 0.0115 0.0141 0.7771 0.0210 clutter 0.1979 0.1911 0.1737 0.0456 0.0033 0.3885 Precision/Correctness 0.833 0.922 0.832 0.820 0.828 0.789 Recall/Completeness 0.916 0.963 0.797 0.866 0.777 0.388 F1 0.872 0.942 0.814 0.843 0.802 0.521 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 0.9379 0.0103 0.0156 0.0222 0.0038 0.0103 building 0.0258 0.9688 0.0033 0.0012 0.0000 0.0009 low_veg 0.0540 0.0050 0.8281 0.0980 0.0001 0.0148 tree 0.0396 0.0029 0.0644 0.8881 0.0025 0.0025 car 0.0965 0.0024 0.0099 0.0092 0.8595 0.0225 clutter 0.1796 0.1663 0.1649 0.0440 0.0032 0.4420 Precision/Correctness 0.866 0.942 0.859 0.844 0.890 0.815 Recall/Completeness 0.938 0.969 0.828 0.888 0.859 0.442 F1 0.901 0.955 0.843 0.865 0.875 0.573 Overall accuracy 0.880 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 0.0385 0.0410 0.0156 0.0069 0.0073 building 0.0177 0.9654 0.0075 0.0052 0.0002 0.0040 low_veg 0.0648 0.0088 0.7950 0.1140 0.0002 0.0171 tree 0.0443 0.0053 0.0754 0.8706 0.0019 0.0024 car 0.1807 0.0061 0.0061 0.0166 0.7861 0.0044 clutter 0.2145 0.1346 0.1993 0.0118 0.0009 0.4390 Precision/Correctness 0.902 0.899 0.734 0.801 0.868 0.820 Recall/Completeness 0.891 0.965 0.795 0.871 0.786 0.439 F1 0.896 0.931 0.764 0.834 0.825 0.572 Overall accuracy 0.859 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 0.9097 0.0340 0.0358 0.0114 0.0028 0.0062 building 0.0132 0.9736 0.0057 0.0038 0.0002 0.0034 low_veg 0.0429 0.0034 0.8306 0.1069 0.0001 0.0161 tree 0.0320 0.0035 0.0660 0.8951 0.0014 0.0021 car 0.0985 0.0061 0.0032 0.0139 0.8735 0.0048 clutter 0.2027 0.0981 0.2010 0.0094 0.0007 0.4882 Precision/Correctness 0.928 0.922 0.757 0.831 0.930 0.843 Recall/Completeness 0.910 0.974 0.831 0.895 0.873 0.488 F1 0.919 0.947 0.792 0.862 0.901 0.618 Overall accuracy 0.885 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 0.9404 0.0142 0.0178 0.0142 0.0037 0.0096 building 0.0354 0.9394 0.0205 0.0011 0.0000 0.0036 low_veg 0.0301 0.0066 0.9186 0.0425 0.0001 0.0020 tree 0.0875 0.0032 0.0807 0.8253 0.0014 0.0020 car 0.1920 0.0003 0.0011 0.0038 0.7910 0.0118 clutter 0.3155 0.0276 0.3454 0.0161 0.0056 0.2898 Precision/Correctness 0.910 0.966 0.885 0.860 0.850 0.526 Recall/Completeness 0.940 0.939 0.919 0.825 0.791 0.290 F1 0.925 0.953 0.901 0.842 0.819 0.374 Overall accuracy 0.906 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 0.9571 0.0108 0.0127 0.0093 0.0016 0.0085 building 0.0306 0.9492 0.0164 0.0006 0.0000 0.0031 low_veg 0.0179 0.0031 0.9414 0.0362 0.0001 0.0013 tree 0.0662 0.0018 0.0697 0.8598 0.0011 0.0015 car 0.1193 0.0003 0.0006 0.0021 0.8628 0.0150 clutter 0.2966 0.0247 0.3549 0.0156 0.0053 0.3029 Precision/Correctness 0.933 0.977 0.905 0.888 0.910 0.546 Recall/Completeness 0.957 0.949 0.941 0.860 0.863 0.303 F1 0.945 0.963 0.923 0.874 0.886 0.390 Overall accuracy 0.927 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 0.9376 0.0172 0.0191 0.0107 0.0122 0.0033 building 0.0358 0.9525 0.0054 0.0014 0.0006 0.0043 low_veg 0.0738 0.0133 0.8213 0.0787 0.0009 0.0120 tree 0.0509 0.0074 0.0648 0.8709 0.0052 0.0007 car 0.1722 0.0038 0.0039 0.0125 0.8043 0.0034 clutter 0.2251 0.2651 0.2472 0.0143 0.0116 0.2368 Precision/Correctness 0.907 0.948 0.857 0.825 0.812 0.545 Recall/Completeness 0.938 0.953 0.821 0.871 0.804 0.237 F1 0.922 0.950 0.839 0.848 0.808 0.330 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: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9574 0.0127 0.0139 0.0077 0.0057 0.0027 building 0.0309 0.9613 0.0029 0.0008 0.0003 0.0038 low_veg 0.0517 0.0061 0.8599 0.0723 0.0003 0.0097 tree 0.0386 0.0059 0.0531 0.8978 0.0039 0.0006 car 0.0962 0.0027 0.0020 0.0082 0.8873 0.0037 clutter 0.1926 0.2876 0.2457 0.0107 0.0119 0.2516 Precision/Correctness 0.932 0.960 0.886 0.854 0.881 0.576 Recall/Completeness 0.957 0.961 0.860 0.898 0.887 0.252 F1 0.945 0.961 0.873 0.875 0.884 0.350 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_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8781 0.0446 0.0279 0.0090 0.0135 0.0269 building 0.0223 0.9106 0.0546 0.0033 0.0000 0.0091 low_veg 0.0700 0.0100 0.8118 0.0639 0.0019 0.0422 tree 0.0561 0.0048 0.1551 0.7727 0.0047 0.0066 car 0.1430 0.0092 0.0079 0.0139 0.8201 0.0058 clutter 0.0860 0.0712 0.1805 0.0065 0.0042 0.6516 Precision/Correctness 0.907 0.899 0.674 0.892 0.797 0.664 Recall/Completeness 0.878 0.911 0.812 0.773 0.820 0.652 F1 0.893 0.905 0.737 0.828 0.808 0.658 Overall accuracy 0.844 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 0.8991 0.0425 0.0201 0.0058 0.0069 0.0257 building 0.0174 0.9201 0.0523 0.0022 0.0000 0.0080 low_veg 0.0480 0.0061 0.8483 0.0548 0.0007 0.0421 tree 0.0444 0.0032 0.1404 0.8022 0.0038 0.0060 car 0.0791 0.0095 0.0027 0.0100 0.8929 0.0058 clutter 0.0735 0.0713 0.1684 0.0048 0.0039 0.6782 Precision/Correctness 0.934 0.910 0.695 0.919 0.862 0.684 Recall/Completeness 0.899 0.920 0.848 0.802 0.893 0.678 F1 0.916 0.915 0.764 0.857 0.877 0.681 Overall accuracy 0.869 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 0.9304 0.0187 0.0113 0.0247 0.0050 0.0100 building 0.0627 0.9215 0.0050 0.0020 0.0001 0.0087 low_veg 0.0642 0.0098 0.7778 0.1364 0.0006 0.0112 tree 0.0381 0.0041 0.1106 0.8458 0.0008 0.0007 car 0.1388 0.0170 0.0018 0.0242 0.8162 0.0020 clutter 0.1530 0.0166 0.1176 0.0023 0.0098 0.7006 Precision/Correctness 0.922 0.944 0.751 0.709 0.762 0.886 Recall/Completeness 0.930 0.921 0.778 0.846 0.816 0.701 F1 0.926 0.932 0.764 0.771 0.788 0.783 Overall accuracy 0.884 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 0.9463 0.0156 0.0084 0.0181 0.0023 0.0093 building 0.0590 0.9279 0.0035 0.0012 0.0001 0.0083 low_veg 0.0467 0.0061 0.8111 0.1260 0.0002 0.0098 tree 0.0263 0.0027 0.1049 0.8652 0.0005 0.0004 car 0.0727 0.0203 0.0008 0.0166 0.8877 0.0018 clutter 0.1422 0.0156 0.1175 0.0019 0.0095 0.7132 Precision/Correctness 0.933 0.954 0.768 0.744 0.818 0.897 Recall/Completeness 0.946 0.928 0.811 0.865 0.888 0.713 F1 0.940 0.941 0.789 0.800 0.852 0.794 Overall accuracy 0.901 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image