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 0.915 0.025 0.035 0.016 0.003 0.006 building 0.026 0.955 0.011 0.004 0.000 0.004 low_veg 0.057 0.012 0.863 0.062 0.000 0.006 tree 0.043 0.006 0.135 0.812 0.002 0.002 car 0.094 0.008 0.003 0.015 0.873 0.006 clutter 0.346 0.143 0.207 0.022 0.011 0.272 Precision/Correctness 0.870 0.927 0.796 0.872 0.901 0.733 Recall/Completeness 0.915 0.955 0.863 0.812 0.873 0.272 F1 0.892 0.941 0.828 0.841 0.887 0.396 Overall accuracy 0.866	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 0.934 0.020 0.028 0.011 0.000 0.005 building 0.022 0.963 0.008 0.003 0.000 0.004 low_veg 0.043 0.007 0.890 0.054 0.000 0.006 tree 0.033 0.004 0.120 0.840 0.002 0.002 car 0.019 0.009 0.001 0.011 0.954 0.007 clutter 0.346 0.129 0.197 0.018 0.011 0.298 Precision/Correctness 0.894 0.944 0.822 0.898 0.939 0.756 Recall/Completeness 0.934 0.963 0.890 0.840 0.954 0.298 F1 0.914 0.954 0.854 0.868 0.946 0.428 Overall accuracy 0.889
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8802 0.0299 0.0599 0.0160 0.0027 0.0114 building 0.0167 0.9455 0.0229 0.0113 0.0005 0.0031 low_veg 0.0419 0.0134 0.8812 0.0540 0.0002 0.0093 tree 0.0351 0.0104 0.1273 0.8232 0.0011 0.0029 car 0.0964 0.0189 0.0019 0.0086 0.8509 0.0232 clutter 0.1554 0.2390 0.1867 0.0303 0.0149 0.3737 Precision/Correctness 0.885 0.903 0.837 0.892 0.878 0.502 Recall/Completeness 0.880 0.945 0.881 0.823 0.851 0.374 F1 0.883 0.924 0.858 0.856 0.864 0.428 Overall accuracy 0.869 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 0.9054 0.0239 0.0492 0.0106 0.0005 0.0103 building 0.0122 0.9565 0.0181 0.0101 0.0005 0.0026 low_veg 0.0342 0.0096 0.9051 0.0426 0.0002 0.0083 tree 0.0270 0.0088 0.1068 0.8544 0.0006 0.0024 car 0.0258 0.0219 0.0007 0.0037 0.9199 0.0279 clutter 0.1131 0.2482 0.1372 0.0244 0.0189 0.4582 Precision/Correctness 0.914 0.926 0.866 0.918 0.926 0.508 Recall/Completeness 0.905 0.956 0.905 0.854 0.920 0.458 F1 0.910 0.941 0.885 0.885 0.923 0.482 Overall accuracy 0.897 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 0.8832 0.0100 0.0874 0.0156 0.0014 0.0024 building 0.0106 0.9700 0.0109 0.0051 0.0000 0.0035 low_veg 0.0128 0.0038 0.9422 0.0401 0.0000 0.0011 tree 0.0153 0.0030 0.1710 0.8095 0.0008 0.0003 car 0.0948 0.0012 0.0026 0.0147 0.8860 0.0006 clutter 0.1246 0.2594 0.2873 0.0726 0.0343 0.2217 Precision/Correctness 0.916 0.929 0.837 0.936 0.890 0.403 Recall/Completeness 0.883 0.970 0.942 0.809 0.886 0.222 F1 0.899 0.949 0.886 0.868 0.888 0.286 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_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9150 0.0067 0.0663 0.0099 0.0001 0.0020 building 0.0062 0.9840 0.0040 0.0028 0.0000 0.0030 low_veg 0.0076 0.0020 0.9573 0.0321 0.0000 0.0010 tree 0.0116 0.0023 0.1547 0.8304 0.0006 0.0003 car 0.0149 0.0004 0.0003 0.0088 0.9753 0.0004 clutter 0.0885 0.3018 0.2397 0.0687 0.0431 0.2582 Precision/Correctness 0.944 0.952 0.856 0.952 0.930 0.391 Recall/Completeness 0.915 0.984 0.957 0.830 0.975 0.258 F1 0.929 0.968 0.904 0.887 0.952 0.311 Overall accuracy 0.905 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 0.8900 0.0215 0.0523 0.0212 0.0034 0.0116 building 0.0136 0.9698 0.0075 0.0041 0.0002 0.0049 low_veg 0.0567 0.0202 0.8540 0.0615 0.0001 0.0075 tree 0.0337 0.0073 0.1117 0.8431 0.0021 0.0022 car 0.0874 0.0067 0.0021 0.0217 0.8813 0.0009 clutter 0.2889 0.1661 0.2799 0.0304 0.0051 0.2295 Precision/Correctness 0.873 0.938 0.806 0.900 0.909 0.461 Recall/Completeness 0.890 0.970 0.854 0.843 0.881 0.230 F1 0.881 0.954 0.829 0.871 0.895 0.306 Overall accuracy 0.875 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 0.9174 0.0151 0.0421 0.0141 0.0003 0.0110 building 0.0102 0.9790 0.0036 0.0028 0.0002 0.0042 low_veg 0.0437 0.0113 0.8887 0.0501 0.0000 0.0062 tree 0.0259 0.0058 0.0953 0.8697 0.0015 0.0018 car 0.0172 0.0074 0.0006 0.0146 0.9595 0.0007 clutter 0.2917 0.1666 0.2435 0.0228 0.0062 0.2693 Precision/Correctness 0.904 0.959 0.839 0.927 0.956 0.468 Recall/Completeness 0.917 0.979 0.889 0.870 0.959 0.269 F1 0.911 0.969 0.863 0.897 0.958 0.342 Overall accuracy 0.904 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 0.8765 0.0305 0.0636 0.0222 0.0033 0.0039 building 0.0136 0.9732 0.0085 0.0033 0.0000 0.0015 low_veg 0.0390 0.0124 0.8971 0.0477 0.0000 0.0037 tree 0.0255 0.0043 0.1871 0.7785 0.0013 0.0033 car 0.0919 0.0026 0.0033 0.0274 0.8721 0.0027 clutter 0.2591 0.1809 0.1955 0.0260 0.0074 0.3311 Precision/Correctness 0.864 0.929 0.778 0.912 0.892 0.768 Recall/Completeness 0.877 0.973 0.897 0.779 0.872 0.331 F1 0.870 0.951 0.833 0.840 0.882 0.463 Overall accuracy 0.862 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 0.9062 0.0247 0.0497 0.0157 0.0004 0.0033 building 0.0106 0.9820 0.0041 0.0021 0.0000 0.0011 low_veg 0.0270 0.0072 0.9244 0.0384 0.0000 0.0031 tree 0.0201 0.0033 0.1695 0.8032 0.0010 0.0028 car 0.0222 0.0014 0.0016 0.0193 0.9526 0.0028 clutter 0.2499 0.1704 0.1544 0.0196 0.0085 0.3972 Precision/Correctness 0.897 0.949 0.805 0.935 0.941 0.799 Recall/Completeness 0.906 0.982 0.924 0.803 0.953 0.397 F1 0.902 0.965 0.861 0.864 0.947 0.531 Overall accuracy 0.889 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 0.9269 0.0139 0.0330 0.0221 0.0021 0.0020 building 0.0165 0.9671 0.0092 0.0048 0.0001 0.0022 low_veg 0.0604 0.0095 0.8236 0.1037 0.0001 0.0027 tree 0.0376 0.0034 0.0578 0.8991 0.0015 0.0006 car 0.0889 0.0022 0.0024 0.0292 0.8752 0.0020 clutter 0.4794 0.1315 0.2209 0.0593 0.0126 0.0964 Precision/Correctness 0.840 0.956 0.771 0.804 0.916 0.777 Recall/Completeness 0.927 0.967 0.824 0.899 0.875 0.096 F1 0.881 0.961 0.796 0.849 0.895 0.171 Overall accuracy 0.864 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 0.9472 0.0090 0.0263 0.0157 0.0002 0.0016 building 0.0112 0.9793 0.0040 0.0035 0.0001 0.0019 low_veg 0.0447 0.0036 0.8563 0.0937 0.0000 0.0017 tree 0.0254 0.0020 0.0440 0.9272 0.0010 0.0004 car 0.0127 0.0018 0.0004 0.0204 0.9627 0.0019 clutter 0.4987 0.1183 0.2107 0.0517 0.0141 0.1064 Precision/Correctness 0.868 0.970 0.810 0.840 0.941 0.815 Recall/Completeness 0.947 0.979 0.856 0.927 0.963 0.106 F1 0.906 0.974 0.833 0.881 0.952 0.188 Overall accuracy 0.892 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 0.8940 0.0271 0.0558 0.0148 0.0039 0.0045 building 0.0192 0.9658 0.0073 0.0045 0.0006 0.0026 low_veg 0.1228 0.0175 0.7950 0.0499 0.0003 0.0144 tree 0.0408 0.0062 0.1678 0.7768 0.0022 0.0062 car 0.0850 0.0125 0.0016 0.0243 0.8741 0.0025 clutter 0.5354 0.0797 0.1375 0.0111 0.0013 0.2350 Precision/Correctness 0.706 0.906 0.750 0.892 0.920 0.789 Recall/Completeness 0.894 0.966 0.795 0.777 0.874 0.235 F1 0.789 0.935 0.772 0.830 0.896 0.362 Overall accuracy 0.798 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 0.9129 0.0229 0.0497 0.0098 0.0006 0.0041 building 0.0148 0.9745 0.0045 0.0033 0.0006 0.0023 low_veg 0.1089 0.0121 0.8221 0.0423 0.0001 0.0144 tree 0.0314 0.0049 0.1494 0.8067 0.0017 0.0058 car 0.0215 0.0127 0.0004 0.0183 0.9444 0.0027 clutter 0.5491 0.0775 0.1207 0.0087 0.0010 0.2431 Precision/Correctness 0.721 0.922 0.773 0.917 0.960 0.805 Recall/Completeness 0.913 0.975 0.822 0.807 0.944 0.243 F1 0.806 0.948 0.797 0.858 0.952 0.374 Overall accuracy 0.818 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 0.8948 0.0394 0.0559 0.0049 0.0025 0.0026 building 0.0540 0.9280 0.0136 0.0021 0.0003 0.0020 low_veg 0.0691 0.0196 0.8695 0.0359 0.0005 0.0053 tree 0.0806 0.0096 0.1563 0.7446 0.0063 0.0025 car 0.1122 0.0196 0.0039 0.0062 0.8562 0.0019 clutter 0.2718 0.2735 0.2608 0.0176 0.0066 0.1697 Precision/Correctness 0.862 0.916 0.737 0.915 0.910 0.640 Recall/Completeness 0.895 0.928 0.870 0.745 0.856 0.170 F1 0.878 0.922 0.798 0.821 0.882 0.268 Overall accuracy 0.860 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 0.9119 0.0331 0.0498 0.0029 0.0003 0.0020 building 0.0521 0.9339 0.0108 0.0015 0.0003 0.0015 low_veg 0.0453 0.0110 0.9102 0.0295 0.0001 0.0039 tree 0.0624 0.0075 0.1352 0.7879 0.0047 0.0022 car 0.0225 0.0236 0.0011 0.0032 0.9472 0.0023 clutter 0.2551 0.2830 0.2587 0.0134 0.0058 0.1840 Precision/Correctness 0.890 0.935 0.764 0.940 0.950 0.677 Recall/Completeness 0.912 0.934 0.910 0.788 0.947 0.184 F1 0.901 0.934 0.831 0.857 0.949 0.289 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 0.9256 0.0255 0.0245 0.0196 0.0020 0.0028 building 0.0163 0.9727 0.0041 0.0032 0.0001 0.0035 low_veg 0.0923 0.0127 0.8078 0.0835 0.0001 0.0035 tree 0.0653 0.0054 0.1135 0.8131 0.0018 0.0008 car 0.1074 0.0058 0.0013 0.0104 0.8676 0.0075 clutter 0.3032 0.2970 0.1646 0.0130 0.0368 0.1854 Precision/Correctness 0.878 0.929 0.843 0.807 0.912 0.672 Recall/Completeness 0.926 0.973 0.808 0.813 0.868 0.185 F1 0.901 0.950 0.825 0.810 0.889 0.291 Overall accuracy 0.877 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 0.9427 0.0199 0.0208 0.0146 0.0002 0.0020 building 0.0126 0.9790 0.0026 0.0025 0.0000 0.0033 low_veg 0.0843 0.0073 0.8304 0.0752 0.0001 0.0027 tree 0.0486 0.0037 0.1010 0.8446 0.0015 0.0007 car 0.0216 0.0068 0.0005 0.0076 0.9549 0.0087 clutter 0.2656 0.2988 0.1588 0.0101 0.0468 0.2199 Precision/Correctness 0.905 0.947 0.869 0.834 0.934 0.694 Recall/Completeness 0.943 0.979 0.830 0.845 0.955 0.220 F1 0.924 0.963 0.849 0.839 0.944 0.334 Overall accuracy 0.902 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 0.9274 0.0168 0.0214 0.0266 0.0016 0.0061 building 0.0147 0.9776 0.0052 0.0020 0.0001 0.0003 low_veg 0.0666 0.0177 0.8212 0.0862 0.0002 0.0081 tree 0.0525 0.0046 0.1102 0.8255 0.0030 0.0041 car 0.1104 0.0024 0.0125 0.0152 0.8399 0.0196 clutter 0.2413 0.2234 0.1761 0.0448 0.0064 0.3081 Precision/Correctness 0.842 0.905 0.811 0.832 0.916 0.835 Recall/Completeness 0.927 0.978 0.821 0.825 0.840 0.308 F1 0.883 0.940 0.816 0.829 0.876 0.450 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 0.9470 0.0129 0.0140 0.0203 0.0001 0.0056 building 0.0129 0.9824 0.0029 0.0014 0.0001 0.0003 low_veg 0.0506 0.0139 0.8464 0.0815 0.0000 0.0075 tree 0.0417 0.0038 0.1013 0.8473 0.0022 0.0038 car 0.0217 0.0016 0.0106 0.0109 0.9335 0.0217 clutter 0.2293 0.1995 0.1677 0.0436 0.0066 0.3532 Precision/Correctness 0.872 0.927 0.835 0.855 0.943 0.853 Recall/Completeness 0.947 0.982 0.846 0.847 0.933 0.353 F1 0.908 0.954 0.841 0.851 0.938 0.499 Overall accuracy 0.877 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 0.8965 0.0479 0.0334 0.0180 0.0023 0.0020 building 0.0187 0.9651 0.0101 0.0050 0.0002 0.0010 low_veg 0.0720 0.0095 0.7980 0.1109 0.0001 0.0095 tree 0.0454 0.0038 0.0831 0.8640 0.0029 0.0008 car 0.1265 0.0075 0.0025 0.0097 0.8488 0.0050 clutter 0.4308 0.1889 0.1207 0.0129 0.0071 0.2395 Precision/Correctness 0.870 0.874 0.770 0.798 0.925 0.865 Recall/Completeness 0.896 0.965 0.798 0.864 0.849 0.240 F1 0.883 0.917 0.784 0.830 0.885 0.375 Overall accuracy 0.848 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 0.9118 0.0438 0.0290 0.0134 0.0006 0.0014 building 0.0143 0.9731 0.0080 0.0037 0.0001 0.0008 low_veg 0.0509 0.0049 0.8307 0.1041 0.0000 0.0093 tree 0.0343 0.0022 0.0733 0.8870 0.0024 0.0008 car 0.0295 0.0084 0.0005 0.0079 0.9478 0.0059 clutter 0.4469 0.1594 0.1092 0.0109 0.0064 0.2672 Precision/Correctness 0.892 0.896 0.795 0.827 0.953 0.885 Recall/Completeness 0.912 0.973 0.831 0.887 0.948 0.267 F1 0.902 0.933 0.813 0.856 0.951 0.411 Overall accuracy 0.872 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 0.9427 0.0252 0.0171 0.0109 0.0019 0.0021 building 0.0367 0.9378 0.0197 0.0011 0.0026 0.0021 low_veg 0.0332 0.0067 0.9265 0.0329 0.0000 0.0008 tree 0.0980 0.0037 0.0986 0.7975 0.0018 0.0005 car 0.0830 0.0009 0.0001 0.0025 0.9049 0.0087 clutter 0.4060 0.0597 0.3056 0.0259 0.0188 0.1839 Precision/Correctness 0.904 0.946 0.882 0.882 0.851 0.697 Recall/Completeness 0.943 0.938 0.926 0.797 0.905 0.184 F1 0.923 0.942 0.904 0.838 0.877 0.291 Overall accuracy 0.905 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 0.9578 0.0219 0.0118 0.0066 0.0004 0.0015 building 0.0317 0.9480 0.0156 0.0006 0.0024 0.0017 low_veg 0.0226 0.0036 0.9453 0.0279 0.0000 0.0006 tree 0.0762 0.0026 0.0868 0.8326 0.0014 0.0004 car 0.0130 0.0006 0.0000 0.0011 0.9763 0.0090 clutter 0.3989 0.0544 0.3074 0.0243 0.0202 0.1947 Precision/Correctness 0.925 0.958 0.903 0.909 0.884 0.739 Recall/Completeness 0.958 0.948 0.945 0.833 0.976 0.195 F1 0.941 0.953 0.924 0.869 0.928 0.308 Overall accuracy 0.924 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 0.9433 0.0196 0.0215 0.0098 0.0045 0.0013 building 0.0428 0.9443 0.0080 0.0015 0.0004 0.0030 low_veg 0.0582 0.0128 0.8518 0.0741 0.0004 0.0027 tree 0.0556 0.0069 0.0769 0.8541 0.0059 0.0008 car 0.0753 0.0032 0.0021 0.0119 0.9049 0.0027 clutter 0.3274 0.2908 0.2427 0.0160 0.0117 0.1114 Precision/Correctness 0.909 0.943 0.846 0.831 0.911 0.564 Recall/Completeness 0.943 0.944 0.852 0.854 0.905 0.111 F1 0.926 0.944 0.849 0.843 0.908 0.186 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_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9598 0.0154 0.0166 0.0067 0.0004 0.0011 building 0.0377 0.9531 0.0052 0.0009 0.0002 0.0029 low_veg 0.0387 0.0059 0.8858 0.0676 0.0000 0.0020 tree 0.0435 0.0054 0.0655 0.8803 0.0047 0.0007 car 0.0114 0.0021 0.0005 0.0081 0.9751 0.0027 clutter 0.3068 0.3231 0.2306 0.0121 0.0115 0.1160 Precision/Correctness 0.930 0.955 0.873 0.861 0.964 0.578 Recall/Completeness 0.960 0.953 0.886 0.880 0.975 0.116 F1 0.945 0.954 0.879 0.870 0.970 0.193 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 0.9266 0.0157 0.0287 0.0093 0.0043 0.0154 building 0.0199 0.9448 0.0291 0.0029 0.0005 0.0028 low_veg 0.0773 0.0061 0.8455 0.0465 0.0009 0.0238 tree 0.0525 0.0053 0.1684 0.7664 0.0038 0.0037 car 0.0801 0.0043 0.0032 0.0132 0.8916 0.0076 clutter 0.2242 0.0441 0.1937 0.0057 0.0177 0.5145 Precision/Correctness 0.897 0.954 0.694 0.910 0.885 0.740 Recall/Completeness 0.927 0.945 0.845 0.766 0.892 0.515 F1 0.912 0.950 0.762 0.832 0.888 0.607 Overall accuracy 0.868 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 0.9448 0.0124 0.0221 0.0058 0.0005 0.0145 building 0.0144 0.9548 0.0260 0.0020 0.0004 0.0024 low_veg 0.0549 0.0033 0.8805 0.0379 0.0001 0.0233 tree 0.0418 0.0040 0.1520 0.7959 0.0029 0.0033 car 0.0137 0.0039 0.0005 0.0092 0.9647 0.0080 clutter 0.2165 0.0422 0.1839 0.0040 0.0175 0.5359 Precision/Correctness 0.921 0.965 0.715 0.936 0.931 0.760 Recall/Completeness 0.945 0.955 0.881 0.796 0.965 0.536 F1 0.933 0.960 0.789 0.860 0.947 0.629 Overall accuracy 0.891 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 0.9383 0.0185 0.0102 0.0225 0.0018 0.0086 building 0.0509 0.9146 0.0058 0.0025 0.0003 0.0258 low_veg 0.0674 0.0081 0.7844 0.1352 0.0002 0.0047 tree 0.0403 0.0034 0.1305 0.8245 0.0009 0.0003 car 0.0742 0.0309 0.0009 0.0225 0.8658 0.0056 clutter 0.2601 0.0211 0.3273 0.0043 0.0133 0.3739 Precision/Correctness 0.910 0.942 0.632 0.711 0.840 0.760 Recall/Completeness 0.938 0.915 0.784 0.825 0.866 0.374 F1 0.924 0.928 0.700 0.764 0.852 0.501 Overall accuracy 0.857 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 0.9526 0.0148 0.0078 0.0162 0.0004 0.0082 building 0.0478 0.9201 0.0046 0.0017 0.0003 0.0256 low_veg 0.0490 0.0045 0.8142 0.1278 0.0000 0.0044 tree 0.0279 0.0021 0.1248 0.8444 0.0005 0.0002 car 0.0122 0.0384 0.0001 0.0160 0.9282 0.0052 clutter 0.2515 0.0194 0.3324 0.0039 0.0127 0.3801 Precision/Correctness 0.921 0.954 0.637 0.744 0.869 0.769 Recall/Completeness 0.953 0.920 0.814 0.844 0.928 0.380 F1 0.936 0.937 0.715 0.791 0.898 0.509 Overall accuracy 0.871 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image