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 0.909 0.023 0.036 0.018 0.003 0.013 building 0.017 0.956 0.013 0.004 0.000 0.010 low_veg 0.040 0.010 0.871 0.065 0.000 0.013 tree 0.037 0.005 0.133 0.819 0.003 0.004 car 0.065 0.007 0.003 0.010 0.898 0.017 clutter 0.249 0.128 0.150 0.024 0.007 0.442 Precision/Correctness 0.901 0.934 0.807 0.867 0.910 0.663 Recall/Completeness 0.909 0.956 0.871 0.819 0.898 0.442 F1 0.905 0.945 0.838 0.842 0.904 0.530 Overall accuracy 0.875	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 0.029 0.013 0.000 0.011 building 0.014 0.963 0.010 0.003 0.000 0.010 low_veg 0.030 0.006 0.896 0.057 0.000 0.012 tree 0.027 0.004 0.118 0.845 0.002 0.003 car 0.011 0.007 0.001 0.006 0.955 0.019 clutter 0.247 0.113 0.139 0.021 0.007 0.474 Precision/Correctness 0.921 0.950 0.830 0.892 0.950 0.686 Recall/Completeness 0.928 0.963 0.896 0.845 0.955 0.474 F1 0.924 0.957 0.861 0.868 0.953 0.561 Overall accuracy 0.897
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8990 0.0229 0.0425 0.0125 0.0024 0.0205 building 0.0266 0.9341 0.0238 0.0076 0.0000 0.0079 low_veg 0.0317 0.0090 0.8957 0.0422 0.0003 0.0211 tree 0.0398 0.0088 0.1619 0.7820 0.0018 0.0057 car 0.0756 0.0112 0.0017 0.0056 0.8536 0.0522 clutter 0.1123 0.2234 0.1309 0.0279 0.0207 0.4848 Precision/Correctness 0.892 0.921 0.832 0.910 0.870 0.386 Recall/Completeness 0.899 0.934 0.896 0.782 0.854 0.485 F1 0.895 0.927 0.862 0.841 0.862 0.430 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 0.9222 0.0170 0.0345 0.0078 0.0003 0.0181 building 0.0235 0.9433 0.0202 0.0066 0.0000 0.0064 low_veg 0.0245 0.0052 0.9174 0.0328 0.0002 0.0200 tree 0.0312 0.0071 0.1396 0.8161 0.0013 0.0047 car 0.0236 0.0126 0.0005 0.0025 0.8986 0.0622 clutter 0.0750 0.2281 0.0981 0.0230 0.0246 0.5513 Precision/Correctness 0.917 0.944 0.857 0.934 0.915 0.370 Recall/Completeness 0.922 0.943 0.917 0.816 0.899 0.551 F1 0.920 0.944 0.886 0.871 0.907 0.443 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: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9011 0.0127 0.0662 0.0108 0.0013 0.0081 building 0.0081 0.9699 0.0097 0.0070 0.0001 0.0051 low_veg 0.0153 0.0036 0.9433 0.0347 0.0000 0.0030 tree 0.0189 0.0032 0.2014 0.7747 0.0009 0.0009 car 0.0699 0.0014 0.0014 0.0078 0.9130 0.0066 clutter 0.1079 0.3215 0.1910 0.0574 0.0335 0.2887 Precision/Correctness 0.906 0.921 0.824 0.943 0.893 0.245 Recall/Completeness 0.901 0.970 0.943 0.775 0.913 0.289 F1 0.903 0.945 0.880 0.851 0.903 0.265 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 0.9260 0.0094 0.0505 0.0069 0.0001 0.0071 building 0.0058 0.9800 0.0049 0.0050 0.0001 0.0042 low_veg 0.0099 0.0018 0.9578 0.0280 0.0000 0.0026 tree 0.0146 0.0024 0.1851 0.7966 0.0006 0.0007 car 0.0158 0.0007 0.0000 0.0046 0.9722 0.0067 clutter 0.0761 0.3715 0.1566 0.0608 0.0408 0.2943 Precision/Correctness 0.932 0.945 0.841 0.956 0.930 0.218 Recall/Completeness 0.926 0.980 0.958 0.797 0.972 0.294 F1 0.929 0.962 0.896 0.869 0.951 0.251 Overall accuracy 0.895 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 0.8871 0.0227 0.0523 0.0166 0.0036 0.0177 building 0.0106 0.9724 0.0080 0.0033 0.0001 0.0057 low_veg 0.0482 0.0233 0.8563 0.0604 0.0001 0.0116 tree 0.0319 0.0077 0.1050 0.8501 0.0028 0.0026 car 0.0603 0.0122 0.0012 0.0150 0.9024 0.0091 clutter 0.2345 0.1858 0.2018 0.0240 0.0029 0.3511 Precision/Correctness 0.889 0.932 0.817 0.909 0.903 0.473 Recall/Completeness 0.887 0.972 0.856 0.850 0.902 0.351 F1 0.888 0.952 0.836 0.878 0.903 0.403 Overall accuracy 0.880 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 0.9141 0.0156 0.0437 0.0104 0.0003 0.0159 building 0.0083 0.9795 0.0053 0.0023 0.0000 0.0046 low_veg 0.0378 0.0144 0.8895 0.0498 0.0000 0.0085 tree 0.0236 0.0061 0.0884 0.8776 0.0021 0.0021 car 0.0072 0.0142 0.0006 0.0097 0.9587 0.0097 clutter 0.2444 0.1866 0.1656 0.0190 0.0036 0.3810 Precision/Correctness 0.916 0.954 0.847 0.933 0.953 0.477 Recall/Completeness 0.914 0.979 0.890 0.878 0.959 0.381 F1 0.915 0.967 0.868 0.904 0.956 0.424 Overall accuracy 0.908 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 0.8777 0.0342 0.0549 0.0190 0.0035 0.0108 building 0.0098 0.9784 0.0057 0.0030 0.0000 0.0031 low_veg 0.0288 0.0120 0.9010 0.0526 0.0000 0.0056 tree 0.0247 0.0041 0.1790 0.7856 0.0014 0.0051 car 0.0694 0.0084 0.0012 0.0203 0.8781 0.0226 clutter 0.2185 0.2177 0.1588 0.0251 0.0068 0.3730 Precision/Correctness 0.886 0.922 0.793 0.911 0.888 0.650 Recall/Completeness 0.878 0.978 0.901 0.786 0.878 0.373 F1 0.882 0.949 0.843 0.844 0.883 0.474 Overall accuracy 0.867 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 0.9073 0.0277 0.0428 0.0129 0.0006 0.0087 building 0.0078 0.9848 0.0029 0.0021 0.0000 0.0024 low_veg 0.0191 0.0066 0.9283 0.0423 0.0000 0.0038 tree 0.0193 0.0030 0.1628 0.8094 0.0011 0.0044 car 0.0180 0.0084 0.0003 0.0135 0.9340 0.0258 clutter 0.2148 0.2146 0.1276 0.0187 0.0076 0.4167 Precision/Correctness 0.914 0.942 0.817 0.934 0.936 0.691 Recall/Completeness 0.907 0.985 0.928 0.809 0.934 0.417 F1 0.911 0.963 0.869 0.867 0.935 0.520 Overall accuracy 0.893 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 0.9262 0.0124 0.0339 0.0210 0.0018 0.0046 building 0.0120 0.9729 0.0066 0.0052 0.0001 0.0032 low_veg 0.0500 0.0086 0.8262 0.1103 0.0002 0.0047 tree 0.0344 0.0026 0.0498 0.9093 0.0028 0.0011 car 0.0644 0.0029 0.0025 0.0156 0.9112 0.0034 clutter 0.4686 0.1223 0.1876 0.0634 0.0075 0.1507 Precision/Correctness 0.852 0.959 0.791 0.802 0.931 0.752 Recall/Completeness 0.926 0.973 0.826 0.909 0.911 0.151 F1 0.888 0.966 0.808 0.852 0.921 0.251 Overall accuracy 0.872 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 0.9454 0.0088 0.0273 0.0149 0.0001 0.0035 building 0.0073 0.9833 0.0031 0.0038 0.0001 0.0025 low_veg 0.0396 0.0031 0.8530 0.1004 0.0000 0.0039 tree 0.0225 0.0014 0.0363 0.9371 0.0019 0.0008 car 0.0073 0.0022 0.0005 0.0098 0.9767 0.0035 clutter 0.4898 0.1088 0.1801 0.0561 0.0079 0.1573 Precision/Correctness 0.876 0.972 0.826 0.836 0.957 0.783 Recall/Completeness 0.945 0.983 0.853 0.937 0.977 0.157 F1 0.909 0.978 0.839 0.884 0.967 0.262 Overall accuracy 0.897 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 0.8915 0.0219 0.0568 0.0134 0.0039 0.0126 building 0.0165 0.9692 0.0058 0.0046 0.0003 0.0037 low_veg 0.0812 0.0148 0.8196 0.0575 0.0003 0.0267 tree 0.0341 0.0056 0.1430 0.8064 0.0025 0.0084 car 0.0644 0.0088 0.0014 0.0146 0.8891 0.0217 clutter 0.4044 0.0639 0.1465 0.0170 0.0009 0.3674 Precision/Correctness 0.765 0.922 0.767 0.886 0.924 0.749 Recall/Completeness 0.892 0.969 0.820 0.806 0.889 0.367 F1 0.823 0.945 0.793 0.844 0.906 0.493 Overall accuracy 0.822 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 0.9102 0.0170 0.0525 0.0091 0.0004 0.0109 building 0.0134 0.9764 0.0037 0.0034 0.0003 0.0029 low_veg 0.0696 0.0094 0.8463 0.0492 0.0001 0.0255 tree 0.0252 0.0041 0.1254 0.8364 0.0019 0.0071 car 0.0219 0.0090 0.0003 0.0102 0.9361 0.0226 clutter 0.4151 0.0607 0.1354 0.0142 0.0006 0.3739 Precision/Correctness 0.778 0.939 0.787 0.910 0.968 0.776 Recall/Completeness 0.910 0.976 0.846 0.836 0.936 0.374 F1 0.839 0.957 0.815 0.872 0.952 0.505 Overall accuracy 0.842 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 0.9063 0.0393 0.0382 0.0053 0.0024 0.0084 building 0.0128 0.9658 0.0105 0.0026 0.0001 0.0082 low_veg 0.0585 0.0253 0.8648 0.0393 0.0004 0.0115 tree 0.0854 0.0093 0.1349 0.7542 0.0090 0.0072 car 0.0774 0.0177 0.0033 0.0018 0.8797 0.0201 clutter 0.2407 0.2527 0.2046 0.0111 0.0059 0.2850 Precision/Correctness 0.903 0.919 0.782 0.911 0.903 0.487 Recall/Completeness 0.906 0.966 0.865 0.754 0.880 0.285 F1 0.905 0.942 0.821 0.825 0.891 0.359 Overall accuracy 0.880 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 0.9227 0.0333 0.0333 0.0034 0.0004 0.0068 building 0.0104 0.9714 0.0088 0.0020 0.0001 0.0073 low_veg 0.0395 0.0168 0.9000 0.0346 0.0000 0.0091 tree 0.0682 0.0074 0.1134 0.7974 0.0071 0.0065 car 0.0099 0.0202 0.0006 0.0004 0.9451 0.0239 clutter 0.2358 0.2493 0.2129 0.0086 0.0050 0.2884 Precision/Correctness 0.929 0.937 0.806 0.932 0.938 0.482 Recall/Completeness 0.923 0.971 0.900 0.797 0.945 0.288 F1 0.926 0.954 0.851 0.859 0.942 0.361 Overall accuracy 0.904 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 0.9102 0.0140 0.0363 0.0250 0.0017 0.0129 building 0.0161 0.9700 0.0055 0.0040 0.0000 0.0044 low_veg 0.0592 0.0097 0.8288 0.0918 0.0000 0.0105 tree 0.0468 0.0048 0.1051 0.8373 0.0030 0.0030 car 0.0847 0.0053 0.0006 0.0063 0.8785 0.0246 clutter 0.2137 0.2707 0.1359 0.0162 0.0228 0.3406 Precision/Correctness 0.908 0.945 0.836 0.789 0.930 0.562 Recall/Completeness 0.910 0.970 0.829 0.837 0.878 0.341 F1 0.909 0.958 0.833 0.812 0.904 0.424 Overall accuracy 0.883 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 0.9297 0.0084 0.0319 0.0193 0.0001 0.0105 building 0.0130 0.9762 0.0038 0.0032 0.0000 0.0038 low_veg 0.0536 0.0052 0.8503 0.0825 0.0000 0.0084 tree 0.0323 0.0033 0.0929 0.8664 0.0025 0.0026 car 0.0115 0.0060 0.0001 0.0042 0.9516 0.0266 clutter 0.1916 0.2677 0.1262 0.0140 0.0286 0.3719 Precision/Correctness 0.931 0.963 0.861 0.817 0.951 0.567 Recall/Completeness 0.930 0.976 0.850 0.866 0.952 0.372 F1 0.930 0.970 0.856 0.841 0.951 0.449 Overall accuracy 0.907 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 0.9264 0.0111 0.0243 0.0251 0.0017 0.0115 building 0.0145 0.9750 0.0053 0.0017 0.0002 0.0033 low_veg 0.0499 0.0064 0.8234 0.1005 0.0003 0.0196 tree 0.0470 0.0045 0.0757 0.8604 0.0047 0.0076 car 0.0786 0.0019 0.0100 0.0088 0.8705 0.0302 clutter 0.1697 0.2046 0.1638 0.0487 0.0072 0.4061 Precision/Correctness 0.873 0.923 0.836 0.826 0.898 0.753 Recall/Completeness 0.926 0.975 0.823 0.860 0.871 0.406 F1 0.899 0.948 0.830 0.843 0.884 0.528 Overall accuracy 0.866 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 0.9456 0.0077 0.0173 0.0193 0.0003 0.0098 building 0.0129 0.9798 0.0031 0.0012 0.0002 0.0028 low_veg 0.0362 0.0030 0.8474 0.0952 0.0000 0.0181 tree 0.0365 0.0035 0.0669 0.8826 0.0036 0.0069 car 0.0089 0.0018 0.0096 0.0058 0.9408 0.0332 clutter 0.1520 0.1793 0.1568 0.0490 0.0074 0.4555 Precision/Correctness 0.901 0.943 0.860 0.848 0.923 0.778 Recall/Completeness 0.946 0.980 0.847 0.883 0.941 0.455 F1 0.923 0.961 0.854 0.865 0.932 0.574 Overall accuracy 0.890 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 0.8861 0.0399 0.0440 0.0180 0.0015 0.0106 building 0.0160 0.9581 0.0109 0.0060 0.0004 0.0086 low_veg 0.0412 0.0055 0.8192 0.1138 0.0001 0.0201 tree 0.0362 0.0030 0.0761 0.8788 0.0033 0.0026 car 0.0935 0.0011 0.0024 0.0090 0.8819 0.0120 clutter 0.2616 0.1431 0.1170 0.0137 0.0034 0.4612 Precision/Correctness 0.910 0.897 0.758 0.796 0.945 0.767 Recall/Completeness 0.886 0.958 0.819 0.879 0.882 0.461 F1 0.898 0.927 0.788 0.835 0.912 0.576 Overall accuracy 0.864 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 0.9004 0.0368 0.0395 0.0137 0.0003 0.0093 building 0.0121 0.9659 0.0094 0.0048 0.0003 0.0074 low_veg 0.0272 0.0015 0.8440 0.1079 0.0000 0.0193 tree 0.0256 0.0013 0.0650 0.9032 0.0027 0.0022 car 0.0204 0.0009 0.0003 0.0072 0.9585 0.0127 clutter 0.2618 0.1099 0.1088 0.0114 0.0030 0.5051 Precision/Correctness 0.930 0.917 0.779 0.824 0.966 0.789 Recall/Completeness 0.900 0.966 0.844 0.903 0.959 0.505 F1 0.915 0.941 0.810 0.862 0.962 0.616 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 0.9316 0.0090 0.0214 0.0234 0.0020 0.0126 building 0.0428 0.9180 0.0210 0.0024 0.0000 0.0158 low_veg 0.0173 0.0024 0.9363 0.0409 0.0000 0.0030 tree 0.0514 0.0015 0.0804 0.8628 0.0015 0.0024 car 0.0392 0.0039 0.0003 0.0029 0.9156 0.0381 clutter 0.2310 0.0245 0.3406 0.0454 0.0100 0.3485 Precision/Correctness 0.932 0.979 0.882 0.837 0.909 0.416 Recall/Completeness 0.932 0.918 0.936 0.863 0.916 0.349 F1 0.932 0.948 0.908 0.850 0.912 0.379 Overall accuracy 0.909 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 0.9498 0.0069 0.0167 0.0161 0.0002 0.0103 building 0.0387 0.9294 0.0162 0.0016 0.0000 0.0140 low_veg 0.0117 0.0011 0.9509 0.0342 0.0000 0.0021 tree 0.0377 0.0010 0.0693 0.8892 0.0011 0.0018 car 0.0038 0.0046 0.0000 0.0012 0.9502 0.0401 clutter 0.2271 0.0201 0.3590 0.0443 0.0092 0.3402 Precision/Correctness 0.945 0.986 0.900 0.870 0.963 0.423 Recall/Completeness 0.950 0.929 0.951 0.889 0.950 0.340 F1 0.947 0.957 0.925 0.880 0.957 0.377 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 0.9417 0.0151 0.0244 0.0099 0.0041 0.0049 building 0.0281 0.9599 0.0076 0.0016 0.0002 0.0026 low_veg 0.0381 0.0125 0.8648 0.0700 0.0005 0.0140 tree 0.0477 0.0051 0.0800 0.8585 0.0073 0.0014 car 0.0477 0.0052 0.0022 0.0054 0.9341 0.0055 clutter 0.1729 0.2175 0.1925 0.0145 0.0143 0.3882 Precision/Correctness 0.938 0.955 0.848 0.839 0.914 0.636 Recall/Completeness 0.942 0.960 0.865 0.858 0.934 0.388 F1 0.940 0.958 0.856 0.849 0.924 0.482 Overall accuracy 0.914 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 0.9577 0.0116 0.0192 0.0072 0.0003 0.0040 building 0.0241 0.9676 0.0051 0.0011 0.0001 0.0020 low_veg 0.0243 0.0083 0.8909 0.0636 0.0000 0.0129 tree 0.0363 0.0039 0.0684 0.8843 0.0059 0.0013 car 0.0055 0.0043 0.0005 0.0028 0.9812 0.0058 clutter 0.1511 0.2352 0.1861 0.0111 0.0133 0.4032 Precision/Correctness 0.955 0.965 0.872 0.866 0.962 0.655 Recall/Completeness 0.958 0.968 0.891 0.884 0.981 0.403 F1 0.956 0.966 0.881 0.875 0.971 0.499 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 0.8875 0.0392 0.0355 0.0141 0.0060 0.0178 building 0.0120 0.9230 0.0548 0.0039 0.0009 0.0055 low_veg 0.0319 0.0020 0.8772 0.0486 0.0011 0.0392 tree 0.0327 0.0037 0.1794 0.7746 0.0047 0.0050 car 0.0393 0.0074 0.0045 0.0126 0.9283 0.0079 clutter 0.1097 0.0388 0.1679 0.0069 0.0057 0.6711 Precision/Correctness 0.943 0.921 0.670 0.894 0.885 0.730 Recall/Completeness 0.888 0.923 0.877 0.775 0.928 0.671 F1 0.914 0.922 0.760 0.830 0.906 0.700 Overall accuracy 0.865 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 0.9073 0.0387 0.0276 0.0102 0.0005 0.0157 building 0.0073 0.9316 0.0526 0.0031 0.0009 0.0045 low_veg 0.0196 0.0006 0.9007 0.0404 0.0001 0.0385 tree 0.0249 0.0026 0.1634 0.8011 0.0038 0.0042 car 0.0060 0.0077 0.0006 0.0090 0.9691 0.0075 clutter 0.1038 0.0383 0.1583 0.0054 0.0052 0.6890 Precision/Correctness 0.959 0.927 0.685 0.920 0.949 0.755 Recall/Completeness 0.907 0.932 0.901 0.801 0.969 0.689 F1 0.932 0.930 0.778 0.857 0.959 0.721 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: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9193 0.0169 0.0135 0.0266 0.0019 0.0217 building 0.0133 0.9007 0.0092 0.0024 0.0000 0.0744 low_veg 0.0457 0.0048 0.7594 0.1793 0.0003 0.0104 tree 0.0196 0.0031 0.0897 0.8862 0.0011 0.0003 car 0.0421 0.0071 0.0010 0.0148 0.9167 0.0183 clutter 0.1499 0.0241 0.0495 0.0020 0.0037 0.7708 Precision/Correctness 0.952 0.947 0.791 0.681 0.901 0.708 Recall/Completeness 0.919 0.901 0.759 0.886 0.917 0.771 F1 0.935 0.923 0.775 0.770 0.909 0.738 Overall accuracy 0.883 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 0.9348 0.0138 0.0109 0.0195 0.0004 0.0205 building 0.0104 0.9065 0.0081 0.0015 0.0000 0.0735 low_veg 0.0353 0.0018 0.7826 0.1714 0.0001 0.0089 tree 0.0120 0.0018 0.0830 0.9024 0.0006 0.0002 car 0.0043 0.0089 0.0002 0.0094 0.9592 0.0180 clutter 0.1432 0.0221 0.0477 0.0015 0.0036 0.7819 Precision/Correctness 0.959 0.957 0.809 0.711 0.946 0.719 Recall/Completeness 0.935 0.906 0.783 0.902 0.959 0.782 F1 0.947 0.931 0.795 0.795 0.953 0.749 Overall accuracy 0.898 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image