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 0.851 0.023 0.044 0.020 0.003 0.059 building 0.015 0.953 0.006 0.008 0.002 0.017 low_veg 0.038 0.013 0.846 0.077 0.001 0.026 tree 0.047 0.012 0.182 0.742 0.004 0.013 car 0.079 0.007 0.005 0.007 0.871 0.031 clutter 0.290 0.109 0.125 0.030 0.020 0.426 Precision/Correctness 0.886 0.930 0.773 0.830 0.851 0.389 Recall/Completeness 0.851 0.953 0.846 0.742 0.871 0.426 F1 0.868 0.941 0.808 0.783 0.861 0.406 Overall accuracy 0.837	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 0.869 0.019 0.036 0.016 0.001 0.059 building 0.012 0.960 0.004 0.007 0.002 0.015 low_veg 0.029 0.008 0.869 0.069 0.000 0.024 tree 0.038 0.011 0.169 0.766 0.003 0.012 car 0.013 0.007 0.002 0.005 0.941 0.032 clutter 0.297 0.093 0.110 0.026 0.019 0.455 Precision/Correctness 0.906 0.946 0.795 0.855 0.877 0.392 Recall/Completeness 0.869 0.960 0.869 0.766 0.941 0.455 F1 0.887 0.953 0.831 0.808 0.908 0.421 Overall accuracy 0.858
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.7300 0.0250 0.0457 0.0098 0.0030 0.1864 building 0.0078 0.9467 0.0079 0.0150 0.0011 0.0214 low_veg 0.0256 0.0137 0.8655 0.0500 0.0004 0.0448 tree 0.0446 0.0205 0.2106 0.7002 0.0019 0.0222 car 0.0908 0.0089 0.0040 0.0028 0.8248 0.0687 clutter 0.0666 0.1719 0.1262 0.0165 0.0400 0.5788 Precision/Correctness 0.891 0.904 0.804 0.887 0.807 0.138 Recall/Completeness 0.730 0.947 0.866 0.700 0.825 0.579 F1 0.803 0.925 0.833 0.783 0.816 0.223 Overall accuracy 0.803 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 0.7483 0.0190 0.0336 0.0065 0.0012 0.1913 building 0.0060 0.9551 0.0055 0.0141 0.0010 0.0185 low_veg 0.0207 0.0089 0.8866 0.0408 0.0003 0.0428 tree 0.0368 0.0188 0.1920 0.7299 0.0014 0.0210 car 0.0346 0.0087 0.0021 0.0012 0.8832 0.0702 clutter 0.0412 0.1661 0.0911 0.0120 0.0441 0.6455 Precision/Correctness 0.916 0.928 0.829 0.910 0.846 0.118 Recall/Completeness 0.748 0.955 0.887 0.730 0.883 0.645 F1 0.824 0.941 0.857 0.810 0.864 0.199 Overall accuracy 0.826 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 0.4819 0.0224 0.0877 0.0101 0.0014 0.3965 building 0.0035 0.9525 0.0080 0.0078 0.0001 0.0281 low_veg 0.0169 0.0052 0.8957 0.0433 0.0001 0.0388 tree 0.0288 0.0107 0.2456 0.6963 0.0010 0.0176 car 0.0563 0.0271 0.0047 0.0050 0.8716 0.0352 clutter 0.0764 0.2816 0.1140 0.0387 0.0594 0.4299 Precision/Correctness 0.804 0.862 0.784 0.925 0.860 0.018 Recall/Completeness 0.482 0.952 0.896 0.696 0.872 0.430 F1 0.603 0.905 0.836 0.794 0.866 0.034 Overall accuracy 0.763 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 0.4840 0.0202 0.0637 0.0061 0.0003 0.4258 building 0.0026 0.9634 0.0048 0.0057 0.0001 0.0235 low_veg 0.0142 0.0030 0.9076 0.0366 0.0001 0.0386 tree 0.0255 0.0099 0.2321 0.7150 0.0008 0.0167 car 0.0053 0.0309 0.0004 0.0026 0.9305 0.0302 clutter 0.0619 0.3233 0.0836 0.0401 0.0701 0.4210 Precision/Correctness 0.823 0.884 0.801 0.940 0.892 0.012 Recall/Completeness 0.484 0.963 0.908 0.715 0.931 0.421 F1 0.610 0.922 0.851 0.812 0.911 0.024 Overall accuracy 0.778 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 0.8295 0.0259 0.0618 0.0224 0.0033 0.0570 building 0.0105 0.9701 0.0047 0.0041 0.0005 0.0100 low_veg 0.0483 0.0266 0.8140 0.0901 0.0005 0.0205 tree 0.0308 0.0120 0.1360 0.8034 0.0039 0.0138 car 0.0740 0.0071 0.0030 0.0094 0.8812 0.0253 clutter 0.2385 0.1840 0.2209 0.0292 0.0162 0.3112 Precision/Correctness 0.882 0.923 0.779 0.868 0.865 0.227 Recall/Completeness 0.829 0.970 0.814 0.803 0.881 0.311 F1 0.855 0.946 0.796 0.834 0.873 0.263 Overall accuracy 0.843 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 0.8577 0.0187 0.0503 0.0160 0.0005 0.0569 building 0.0085 0.9763 0.0031 0.0031 0.0004 0.0085 low_veg 0.0404 0.0157 0.8451 0.0804 0.0003 0.0182 tree 0.0244 0.0102 0.1212 0.8274 0.0032 0.0135 car 0.0122 0.0073 0.0007 0.0052 0.9502 0.0243 clutter 0.2478 0.1807 0.1916 0.0223 0.0176 0.3401 Precision/Correctness 0.908 0.947 0.807 0.893 0.905 0.207 Recall/Completeness 0.858 0.976 0.845 0.827 0.950 0.340 F1 0.882 0.962 0.826 0.859 0.927 0.257 Overall accuracy 0.870 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 0.7863 0.0426 0.0695 0.0240 0.0044 0.0731 building 0.0086 0.9737 0.0032 0.0050 0.0002 0.0093 low_veg 0.0272 0.0156 0.8611 0.0808 0.0002 0.0150 tree 0.0313 0.0175 0.2584 0.6719 0.0023 0.0186 car 0.0663 0.0177 0.0043 0.0136 0.8684 0.0296 clutter 0.1879 0.1738 0.1638 0.0275 0.0161 0.4308 Precision/Correctness 0.874 0.902 0.728 0.857 0.831 0.328 Recall/Completeness 0.786 0.974 0.861 0.672 0.868 0.431 F1 0.828 0.937 0.789 0.753 0.849 0.373 Overall accuracy 0.808 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 0.8135 0.0364 0.0566 0.0176 0.0017 0.0743 building 0.0070 0.9800 0.0011 0.0041 0.0002 0.0075 low_veg 0.0198 0.0092 0.8875 0.0704 0.0002 0.0129 tree 0.0270 0.0161 0.2476 0.6894 0.0020 0.0178 car 0.0095 0.0180 0.0017 0.0080 0.9353 0.0275 clutter 0.1841 0.1625 0.1328 0.0206 0.0171 0.4830 Precision/Correctness 0.899 0.924 0.746 0.883 0.864 0.326 Recall/Completeness 0.813 0.980 0.887 0.689 0.935 0.483 F1 0.854 0.951 0.811 0.774 0.898 0.389 Overall accuracy 0.832 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 0.8912 0.0239 0.0368 0.0301 0.0016 0.0165 building 0.0110 0.9666 0.0055 0.0079 0.0010 0.0081 low_veg 0.0402 0.0203 0.7815 0.1503 0.0005 0.0073 tree 0.0452 0.0065 0.0690 0.8698 0.0036 0.0058 car 0.0852 0.0070 0.0056 0.0107 0.8748 0.0167 clutter 0.3991 0.1276 0.1634 0.0674 0.0160 0.2265 Precision/Correctness 0.859 0.942 0.775 0.744 0.887 0.609 Recall/Completeness 0.891 0.967 0.782 0.870 0.875 0.226 F1 0.875 0.954 0.778 0.802 0.881 0.330 Overall accuracy 0.852 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 0.9104 0.0200 0.0301 0.0238 0.0004 0.0153 building 0.0071 0.9756 0.0027 0.0066 0.0007 0.0073 low_veg 0.0314 0.0136 0.8077 0.1412 0.0002 0.0058 tree 0.0342 0.0045 0.0577 0.8956 0.0028 0.0052 car 0.0116 0.0056 0.0020 0.0060 0.9574 0.0174 clutter 0.4116 0.1119 0.1530 0.0600 0.0160 0.2475 Precision/Correctness 0.884 0.957 0.807 0.776 0.908 0.626 Recall/Completeness 0.910 0.976 0.808 0.896 0.957 0.247 F1 0.897 0.966 0.807 0.831 0.932 0.355 Overall accuracy 0.878 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 0.8538 0.0203 0.0661 0.0177 0.0031 0.0390 building 0.0107 0.9680 0.0025 0.0050 0.0063 0.0074 low_veg 0.0829 0.0187 0.7570 0.0855 0.0010 0.0549 tree 0.0459 0.0099 0.1500 0.7749 0.0043 0.0149 car 0.0773 0.0071 0.0049 0.0124 0.8692 0.0291 clutter 0.5577 0.0430 0.1021 0.0158 0.0041 0.2773 Precision/Correctness 0.715 0.925 0.757 0.844 0.850 0.499 Recall/Completeness 0.854 0.968 0.757 0.775 0.869 0.277 F1 0.778 0.946 0.757 0.808 0.859 0.357 Overall accuracy 0.783 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 0.8711 0.0154 0.0611 0.0137 0.0005 0.0383 building 0.0083 0.9746 0.0011 0.0038 0.0061 0.0060 low_veg 0.0725 0.0128 0.7806 0.0783 0.0006 0.0553 tree 0.0378 0.0081 0.1330 0.8036 0.0036 0.0138 car 0.0198 0.0072 0.0029 0.0090 0.9309 0.0302 clutter 0.5790 0.0391 0.0878 0.0127 0.0031 0.2782 Precision/Correctness 0.724 0.943 0.779 0.868 0.873 0.514 Recall/Completeness 0.871 0.975 0.781 0.804 0.931 0.278 F1 0.791 0.959 0.780 0.834 0.901 0.361 Overall accuracy 0.801 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 0.8753 0.0246 0.0581 0.0093 0.0031 0.0296 building 0.0053 0.9752 0.0060 0.0094 0.0004 0.0036 low_veg 0.0464 0.0262 0.8572 0.0562 0.0020 0.0119 tree 0.0962 0.0130 0.1665 0.7016 0.0119 0.0109 car 0.0831 0.0041 0.0067 0.0027 0.8697 0.0338 clutter 0.3025 0.2306 0.1108 0.0254 0.0187 0.3120 Precision/Correctness 0.902 0.935 0.752 0.848 0.853 0.370 Recall/Completeness 0.875 0.975 0.857 0.702 0.870 0.312 F1 0.889 0.955 0.801 0.768 0.861 0.338 Overall accuracy 0.865 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 0.8915 0.0187 0.0530 0.0070 0.0010 0.0288 building 0.0035 0.9809 0.0045 0.0085 0.0002 0.0024 low_veg 0.0303 0.0171 0.8913 0.0499 0.0017 0.0097 tree 0.0832 0.0112 0.1463 0.7384 0.0105 0.0104 car 0.0134 0.0040 0.0012 0.0013 0.9435 0.0366 clutter 0.3179 0.2319 0.0876 0.0226 0.0153 0.3245 Precision/Correctness 0.926 0.952 0.776 0.874 0.883 0.353 Recall/Completeness 0.891 0.981 0.891 0.738 0.944 0.325 F1 0.909 0.966 0.829 0.801 0.912 0.338 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_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8719 0.0266 0.0467 0.0246 0.0021 0.0280 building 0.0102 0.9513 0.0047 0.0066 0.0006 0.0266 low_veg 0.0421 0.0125 0.8468 0.0806 0.0008 0.0171 tree 0.0573 0.0074 0.1829 0.7434 0.0042 0.0047 car 0.1012 0.0037 0.0009 0.0030 0.8643 0.0269 clutter 0.1759 0.2253 0.1368 0.0332 0.1045 0.3242 Precision/Correctness 0.917 0.933 0.787 0.775 0.818 0.327 Recall/Completeness 0.872 0.951 0.847 0.743 0.864 0.324 F1 0.894 0.942 0.816 0.759 0.840 0.326 Overall accuracy 0.856 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 0.8920 0.0214 0.0410 0.0196 0.0005 0.0255 building 0.0074 0.9573 0.0033 0.0057 0.0005 0.0257 low_veg 0.0372 0.0077 0.8666 0.0735 0.0005 0.0144 tree 0.0440 0.0058 0.1694 0.7730 0.0037 0.0041 car 0.0170 0.0043 0.0001 0.0014 0.9472 0.0298 clutter 0.1514 0.2108 0.1252 0.0330 0.1268 0.3528 Precision/Correctness 0.940 0.951 0.812 0.802 0.833 0.306 Recall/Completeness 0.892 0.957 0.867 0.773 0.947 0.353 F1 0.916 0.954 0.839 0.787 0.887 0.328 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8848 0.0116 0.0370 0.0293 0.0014 0.0358 building 0.0161 0.9628 0.0040 0.0026 0.0004 0.0141 low_veg 0.0535 0.0074 0.8083 0.1036 0.0006 0.0265 tree 0.0610 0.0117 0.1423 0.7594 0.0069 0.0187 car 0.0946 0.0126 0.0098 0.0064 0.8472 0.0294 clutter 0.1691 0.1736 0.1402 0.0488 0.0145 0.4539 Precision/Correctness 0.855 0.924 0.783 0.798 0.852 0.599 Recall/Completeness 0.885 0.963 0.808 0.759 0.847 0.454 F1 0.870 0.943 0.795 0.778 0.850 0.517 Overall accuracy 0.834 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 0.9028 0.0091 0.0295 0.0245 0.0003 0.0338 building 0.0143 0.9681 0.0023 0.0021 0.0003 0.0130 low_veg 0.0404 0.0038 0.8308 0.0999 0.0003 0.0249 tree 0.0513 0.0105 0.1341 0.7809 0.0059 0.0172 car 0.0137 0.0135 0.0077 0.0037 0.9316 0.0297 clutter 0.1545 0.1447 0.1309 0.0490 0.0141 0.5068 Precision/Correctness 0.883 0.944 0.804 0.819 0.871 0.620 Recall/Completeness 0.903 0.968 0.831 0.781 0.932 0.507 F1 0.893 0.956 0.817 0.799 0.900 0.558 Overall accuracy 0.857 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 0.8648 0.0076 0.0449 0.0260 0.0018 0.0550 building 0.0128 0.9529 0.0078 0.0119 0.0024 0.0122 low_veg 0.0439 0.0091 0.7814 0.1394 0.0003 0.0259 tree 0.0470 0.0082 0.1201 0.8138 0.0052 0.0057 car 0.1076 0.0023 0.0045 0.0100 0.8459 0.0297 clutter 0.3410 0.1086 0.0930 0.0329 0.0112 0.4133 Precision/Correctness 0.892 0.948 0.734 0.725 0.890 0.490 Recall/Completeness 0.865 0.953 0.781 0.814 0.846 0.413 F1 0.878 0.950 0.757 0.767 0.868 0.449 Overall accuracy 0.837 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 0.8780 0.0044 0.0401 0.0215 0.0008 0.0551 building 0.0092 0.9621 0.0065 0.0098 0.0021 0.0102 low_veg 0.0281 0.0047 0.8070 0.1351 0.0003 0.0249 tree 0.0370 0.0066 0.1104 0.8364 0.0047 0.0049 car 0.0196 0.0022 0.0008 0.0081 0.9379 0.0313 clutter 0.3558 0.0750 0.0832 0.0319 0.0104 0.4436 Precision/Correctness 0.912 0.968 0.755 0.752 0.903 0.496 Recall/Completeness 0.878 0.962 0.807 0.836 0.938 0.444 F1 0.895 0.965 0.780 0.792 0.920 0.468 Overall accuracy 0.858 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 0.9110 0.0316 0.0246 0.0136 0.0015 0.0177 building 0.0293 0.9336 0.0141 0.0118 0.0010 0.0103 low_veg 0.0190 0.0070 0.9396 0.0324 0.0002 0.0019 tree 0.1148 0.0062 0.1754 0.6973 0.0021 0.0042 car 0.0672 0.0104 0.0008 0.0008 0.7946 0.1262 clutter 0.2965 0.0631 0.2062 0.0279 0.0221 0.3843 Precision/Correctness 0.914 0.935 0.855 0.837 0.864 0.404 Recall/Completeness 0.911 0.934 0.940 0.697 0.795 0.384 F1 0.913 0.934 0.895 0.761 0.828 0.394 Overall accuracy 0.886 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 0.9248 0.0288 0.0196 0.0099 0.0004 0.0164 building 0.0257 0.9416 0.0111 0.0113 0.0010 0.0093 low_veg 0.0111 0.0042 0.9553 0.0280 0.0001 0.0012 tree 0.0953 0.0050 0.1639 0.7302 0.0017 0.0038 car 0.0109 0.0131 0.0006 0.0002 0.8416 0.1335 clutter 0.2936 0.0571 0.1925 0.0261 0.0188 0.4119 Precision/Correctness 0.932 0.945 0.873 0.861 0.898 0.415 Recall/Completeness 0.925 0.942 0.955 0.730 0.842 0.412 F1 0.928 0.943 0.913 0.790 0.869 0.413 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9258 0.0055 0.0332 0.0123 0.0036 0.0196 building 0.0369 0.9194 0.0044 0.0029 0.0069 0.0295 low_veg 0.0423 0.0097 0.8719 0.0661 0.0007 0.0094 tree 0.0573 0.0099 0.1747 0.7436 0.0080 0.0064 car 0.0555 0.0040 0.0060 0.0071 0.9111 0.0163 clutter 0.3262 0.1824 0.2043 0.0496 0.0202 0.2172 Precision/Correctness 0.918 0.969 0.789 0.804 0.856 0.209 Recall/Completeness 0.926 0.919 0.872 0.744 0.911 0.217 F1 0.922 0.943 0.829 0.773 0.883 0.213 Overall accuracy 0.881 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 0.9405 0.0027 0.0280 0.0099 0.0002 0.0186 building 0.0328 0.9273 0.0023 0.0024 0.0066 0.0286 low_veg 0.0276 0.0058 0.8980 0.0606 0.0002 0.0079 tree 0.0477 0.0082 0.1617 0.7697 0.0066 0.0060 car 0.0043 0.0035 0.0041 0.0047 0.9678 0.0155 clutter 0.3256 0.1986 0.1991 0.0480 0.0190 0.2096 Precision/Correctness 0.935 0.977 0.811 0.831 0.882 0.191 Recall/Completeness 0.941 0.927 0.898 0.770 0.968 0.210 F1 0.938 0.951 0.852 0.799 0.923 0.200 Overall accuracy 0.899 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 0.8752 0.0098 0.0401 0.0256 0.0038 0.0455 building 0.0095 0.9399 0.0105 0.0162 0.0014 0.0225 low_veg 0.0324 0.0040 0.8435 0.0835 0.0009 0.0357 tree 0.0281 0.0103 0.1970 0.7536 0.0056 0.0053 car 0.0689 0.0019 0.0056 0.0094 0.8896 0.0247 clutter 0.1454 0.0184 0.1376 0.0235 0.0318 0.6434 Precision/Correctness 0.937 0.969 0.689 0.810 0.854 0.568 Recall/Completeness 0.875 0.940 0.843 0.754 0.890 0.643 F1 0.905 0.954 0.758 0.781 0.871 0.603 Overall accuracy 0.853 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 0.8931 0.0074 0.0331 0.0214 0.0004 0.0446 building 0.0064 0.9490 0.0082 0.0153 0.0012 0.0199 low_veg 0.0180 0.0019 0.8698 0.0758 0.0001 0.0343 tree 0.0210 0.0090 0.1803 0.7804 0.0045 0.0047 car 0.0104 0.0018 0.0010 0.0064 0.9553 0.0252 clutter 0.1416 0.0160 0.1307 0.0198 0.0302 0.6618 Precision/Correctness 0.955 0.977 0.707 0.836 0.888 0.585 Recall/Completeness 0.893 0.949 0.870 0.780 0.955 0.662 F1 0.923 0.963 0.780 0.807 0.921 0.621 Overall accuracy 0.874 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 0.8576 0.0471 0.0170 0.0257 0.0022 0.0504 building 0.0238 0.9228 0.0037 0.0022 0.0044 0.0432 low_veg 0.0439 0.0178 0.8026 0.1201 0.0005 0.0152 tree 0.0304 0.0151 0.1635 0.7817 0.0022 0.0070 car 0.0549 0.0052 0.0023 0.0079 0.9201 0.0095 clutter 0.2025 0.0259 0.0454 0.0019 0.0109 0.7133 Precision/Correctness 0.932 0.878 0.756 0.703 0.794 0.636 Recall/Completeness 0.858 0.923 0.803 0.782 0.920 0.713 F1 0.893 0.900 0.779 0.740 0.853 0.672 Overall accuracy 0.849 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 0.8698 0.0437 0.0142 0.0206 0.0008 0.0509 building 0.0214 0.9270 0.0028 0.0017 0.0040 0.0431 low_veg 0.0315 0.0135 0.8257 0.1141 0.0003 0.0148 tree 0.0173 0.0124 0.1528 0.8088 0.0016 0.0072 car 0.0051 0.0043 0.0006 0.0044 0.9773 0.0084 clutter 0.1964 0.0235 0.0432 0.0016 0.0097 0.7255 Precision/Correctness 0.941 0.890 0.776 0.730 0.816 0.641 Recall/Completeness 0.870 0.927 0.826 0.809 0.977 0.726 F1 0.904 0.908 0.800 0.768 0.889 0.681 Overall accuracy 0.862 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image