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 0.917 0.015 0.037 0.018 0.002 0.012 building 0.016 0.967 0.006 0.004 0.000 0.007 low_veg 0.043 0.009 0.868 0.069 0.000 0.011 tree 0.039 0.006 0.132 0.816 0.002 0.004 car 0.094 0.003 0.003 0.014 0.863 0.023 clutter 0.309 0.095 0.149 0.023 0.013 0.411 Precision/Correctness 0.891 0.951 0.812 0.862 0.905 0.677 Recall/Completeness 0.917 0.967 0.868 0.816 0.863 0.411 F1 0.904 0.959 0.839 0.838 0.884 0.511 Overall accuracy 0.877	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 0.011 0.030 0.013 0.000 0.011 building 0.012 0.975 0.003 0.003 0.000 0.006 low_veg 0.031 0.005 0.893 0.061 0.000 0.010 tree 0.029 0.005 0.117 0.842 0.002 0.004 car 0.018 0.003 0.001 0.009 0.943 0.025 clutter 0.309 0.080 0.137 0.019 0.013 0.441 Precision/Correctness 0.913 0.965 0.835 0.888 0.936 0.697 Recall/Completeness 0.934 0.975 0.893 0.842 0.943 0.441 F1 0.924 0.970 0.863 0.864 0.939 0.540 Overall accuracy 0.900
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8942 0.0128 0.0515 0.0110 0.0019 0.0286 building 0.0162 0.9562 0.0143 0.0065 0.0000 0.0068 low_veg 0.0247 0.0070 0.9122 0.0415 0.0002 0.0145 tree 0.0426 0.0087 0.1469 0.7940 0.0013 0.0066 car 0.1136 0.0021 0.0048 0.0089 0.7948 0.0758 clutter 0.1433 0.1796 0.1433 0.0230 0.0280 0.4828 Precision/Correctness 0.900 0.943 0.841 0.916 0.870 0.381 Recall/Completeness 0.894 0.956 0.912 0.794 0.795 0.483 F1 0.897 0.949 0.875 0.850 0.831 0.426 Overall accuracy 0.879 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 0.9158 0.0080 0.0421 0.0070 0.0002 0.0269 building 0.0123 0.9663 0.0103 0.0054 0.0000 0.0056 low_veg 0.0175 0.0036 0.9337 0.0319 0.0001 0.0131 tree 0.0337 0.0071 0.1261 0.8266 0.0009 0.0056 car 0.0443 0.0024 0.0028 0.0039 0.8577 0.0889 clutter 0.1123 0.1784 0.1060 0.0179 0.0335 0.5519 Precision/Correctness 0.927 0.962 0.867 0.938 0.912 0.363 Recall/Completeness 0.916 0.966 0.934 0.827 0.858 0.552 F1 0.922 0.964 0.899 0.879 0.884 0.438 Overall accuracy 0.904 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 0.8690 0.0052 0.1033 0.0103 0.0011 0.0111 building 0.0060 0.9734 0.0100 0.0060 0.0000 0.0045 low_veg 0.0101 0.0030 0.9390 0.0461 0.0000 0.0018 tree 0.0191 0.0042 0.2017 0.7715 0.0007 0.0028 car 0.0947 0.0068 0.0034 0.0071 0.8725 0.0155 clutter 0.0665 0.3056 0.1804 0.0632 0.0036 0.3807 Precision/Correctness 0.917 0.935 0.813 0.928 0.922 0.262 Recall/Completeness 0.869 0.973 0.939 0.772 0.873 0.381 F1 0.892 0.954 0.872 0.842 0.896 0.310 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_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8988 0.0024 0.0813 0.0068 0.0001 0.0106 building 0.0031 0.9843 0.0047 0.0042 0.0000 0.0036 low_veg 0.0062 0.0012 0.9523 0.0389 0.0000 0.0014 tree 0.0152 0.0034 0.1859 0.7923 0.0005 0.0027 car 0.0244 0.0066 0.0004 0.0030 0.9499 0.0157 clutter 0.0378 0.3571 0.1289 0.0668 0.0036 0.4058 Precision/Correctness 0.940 0.957 0.832 0.942 0.956 0.232 Recall/Completeness 0.899 0.984 0.952 0.792 0.950 0.406 F1 0.919 0.971 0.888 0.861 0.953 0.295 Overall accuracy 0.888 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 0.8924 0.0164 0.0595 0.0186 0.0028 0.0104 building 0.0102 0.9780 0.0044 0.0037 0.0000 0.0036 low_veg 0.0504 0.0181 0.8636 0.0612 0.0001 0.0066 tree 0.0303 0.0083 0.1102 0.8448 0.0028 0.0037 car 0.0890 0.0053 0.0013 0.0185 0.8590 0.0268 clutter 0.3191 0.1273 0.2413 0.0218 0.0083 0.2821 Precision/Correctness 0.881 0.948 0.808 0.905 0.904 0.514 Recall/Completeness 0.892 0.978 0.864 0.845 0.859 0.282 F1 0.887 0.963 0.835 0.874 0.881 0.364 Overall accuracy 0.881 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 0.9180 0.0102 0.0502 0.0119 0.0002 0.0095 building 0.0072 0.9854 0.0018 0.0026 0.0000 0.0029 low_veg 0.0387 0.0092 0.8964 0.0508 0.0001 0.0048 tree 0.0222 0.0067 0.0942 0.8716 0.0021 0.0031 car 0.0133 0.0060 0.0001 0.0126 0.9371 0.0309 clutter 0.3331 0.1198 0.2159 0.0141 0.0086 0.3085 Precision/Correctness 0.912 0.968 0.837 0.930 0.947 0.514 Recall/Completeness 0.918 0.985 0.896 0.872 0.937 0.308 F1 0.915 0.977 0.866 0.900 0.942 0.385 Overall accuracy 0.909 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 0.8715 0.0168 0.0779 0.0208 0.0030 0.0101 building 0.0106 0.9774 0.0044 0.0030 0.0000 0.0046 low_veg 0.0259 0.0082 0.9028 0.0577 0.0000 0.0053 tree 0.0237 0.0057 0.1875 0.7755 0.0016 0.0060 car 0.0885 0.0097 0.0018 0.0313 0.8386 0.0300 clutter 0.2344 0.1417 0.1780 0.0249 0.0054 0.4156 Precision/Correctness 0.886 0.949 0.775 0.902 0.892 0.660 Recall/Completeness 0.872 0.977 0.903 0.776 0.839 0.416 F1 0.879 0.963 0.834 0.834 0.865 0.510 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 0.9002 0.0110 0.0655 0.0141 0.0004 0.0087 building 0.0082 0.9847 0.0013 0.0020 0.0000 0.0038 low_veg 0.0161 0.0034 0.9290 0.0475 0.0000 0.0039 tree 0.0188 0.0047 0.1717 0.7982 0.0014 0.0052 car 0.0208 0.0105 0.0004 0.0230 0.9107 0.0347 clutter 0.2225 0.1318 0.1514 0.0166 0.0064 0.4714 Precision/Correctness 0.916 0.966 0.798 0.926 0.934 0.690 Recall/Completeness 0.900 0.985 0.929 0.798 0.911 0.471 F1 0.908 0.975 0.858 0.858 0.922 0.560 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_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9235 0.0098 0.0368 0.0233 0.0016 0.0050 building 0.0117 0.9716 0.0066 0.0049 0.0000 0.0052 low_veg 0.0504 0.0078 0.8350 0.1036 0.0001 0.0030 tree 0.0347 0.0032 0.0554 0.9038 0.0020 0.0009 car 0.0974 0.0007 0.0040 0.0197 0.8704 0.0078 clutter 0.5479 0.0997 0.1577 0.0660 0.0119 0.1169 Precision/Correctness 0.836 0.966 0.795 0.802 0.923 0.666 Recall/Completeness 0.924 0.972 0.835 0.904 0.870 0.117 F1 0.878 0.969 0.815 0.850 0.896 0.199 Overall accuracy 0.868 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 0.9409 0.0070 0.0305 0.0169 0.0004 0.0043 building 0.0065 0.9822 0.0028 0.0036 0.0000 0.0049 low_veg 0.0374 0.0037 0.8634 0.0934 0.0000 0.0021 tree 0.0225 0.0020 0.0418 0.9318 0.0013 0.0005 car 0.0156 0.0002 0.0010 0.0123 0.9628 0.0081 clutter 0.5778 0.0864 0.1434 0.0594 0.0128 0.1203 Precision/Correctness 0.862 0.977 0.831 0.836 0.941 0.675 Recall/Completeness 0.941 0.982 0.863 0.932 0.963 0.120 F1 0.900 0.979 0.847 0.882 0.952 0.204 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8800 0.0172 0.0627 0.0159 0.0029 0.0213 building 0.0078 0.9783 0.0043 0.0042 0.0000 0.0053 low_veg 0.0687 0.0135 0.8273 0.0624 0.0003 0.0278 tree 0.0290 0.0058 0.1405 0.8138 0.0022 0.0086 car 0.0820 0.0025 0.0023 0.0241 0.8592 0.0298 clutter 0.5175 0.0364 0.1594 0.0120 0.0012 0.2736 Precision/Correctness 0.747 0.942 0.763 0.880 0.936 0.641 Recall/Completeness 0.880 0.978 0.827 0.814 0.859 0.274 F1 0.808 0.960 0.794 0.845 0.896 0.383 Overall accuracy 0.815 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 0.8979 0.0123 0.0580 0.0107 0.0003 0.0207 building 0.0054 0.9850 0.0020 0.0031 0.0000 0.0045 low_veg 0.0565 0.0083 0.8538 0.0538 0.0001 0.0275 tree 0.0208 0.0044 0.1230 0.8426 0.0017 0.0076 car 0.0199 0.0020 0.0008 0.0169 0.9288 0.0316 clutter 0.5343 0.0328 0.1479 0.0094 0.0008 0.2748 Precision/Correctness 0.759 0.958 0.782 0.905 0.971 0.658 Recall/Completeness 0.898 0.985 0.854 0.843 0.929 0.275 F1 0.823 0.971 0.816 0.873 0.950 0.388 Overall accuracy 0.834 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 0.9202 0.0230 0.0421 0.0069 0.0024 0.0053 building 0.0043 0.9872 0.0043 0.0016 0.0000 0.0026 low_veg 0.0538 0.0220 0.8749 0.0406 0.0005 0.0081 tree 0.0685 0.0097 0.1339 0.7734 0.0075 0.0070 car 0.1066 0.0033 0.0043 0.0048 0.8660 0.0151 clutter 0.3698 0.2029 0.1211 0.0148 0.0107 0.2807 Precision/Correctness 0.907 0.943 0.797 0.908 0.905 0.609 Recall/Completeness 0.920 0.987 0.875 0.773 0.866 0.281 F1 0.914 0.964 0.834 0.835 0.885 0.384 Overall accuracy 0.895 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 0.9370 0.0174 0.0367 0.0047 0.0004 0.0039 building 0.0021 0.9930 0.0019 0.0010 0.0000 0.0021 low_veg 0.0345 0.0140 0.9113 0.0336 0.0001 0.0066 tree 0.0517 0.0077 0.1135 0.8149 0.0059 0.0064 car 0.0239 0.0034 0.0008 0.0023 0.9525 0.0170 clutter 0.3933 0.2009 0.1062 0.0106 0.0086 0.2805 Precision/Correctness 0.934 0.959 0.825 0.933 0.943 0.616 Recall/Completeness 0.937 0.993 0.911 0.815 0.953 0.280 F1 0.936 0.975 0.866 0.870 0.948 0.385 Overall accuracy 0.920 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 0.9212 0.0125 0.0331 0.0210 0.0019 0.0103 building 0.0114 0.9770 0.0038 0.0035 0.0002 0.0041 low_veg 0.0551 0.0077 0.8504 0.0817 0.0001 0.0050 tree 0.0576 0.0053 0.1117 0.8211 0.0023 0.0020 car 0.1065 0.0008 0.0011 0.0078 0.8625 0.0212 clutter 0.2599 0.2138 0.1318 0.0193 0.0557 0.3195 Precision/Correctness 0.906 0.955 0.843 0.806 0.888 0.618 Recall/Completeness 0.921 0.977 0.850 0.821 0.862 0.320 F1 0.913 0.966 0.847 0.814 0.875 0.421 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_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9399 0.0075 0.0280 0.0156 0.0002 0.0089 building 0.0074 0.9840 0.0020 0.0027 0.0001 0.0038 low_veg 0.0484 0.0037 0.8713 0.0730 0.0000 0.0035 tree 0.0426 0.0037 0.0981 0.8521 0.0018 0.0017 car 0.0194 0.0007 0.0002 0.0056 0.9507 0.0234 clutter 0.2360 0.1979 0.1211 0.0175 0.0682 0.3594 Precision/Correctness 0.932 0.971 0.870 0.835 0.909 0.622 Recall/Completeness 0.940 0.984 0.871 0.852 0.951 0.359 F1 0.936 0.978 0.871 0.843 0.929 0.456 Overall accuracy 0.915 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 0.9209 0.0048 0.0228 0.0291 0.0014 0.0211 building 0.0170 0.9719 0.0044 0.0022 0.0001 0.0044 low_veg 0.0672 0.0086 0.7871 0.1200 0.0002 0.0169 tree 0.0497 0.0063 0.0696 0.8662 0.0035 0.0046 car 0.1060 0.0104 0.0097 0.0158 0.8300 0.0281 clutter 0.2129 0.1693 0.1730 0.0424 0.0048 0.3975 Precision/Correctness 0.847 0.935 0.834 0.805 0.920 0.717 Recall/Completeness 0.921 0.972 0.787 0.866 0.830 0.398 F1 0.882 0.953 0.810 0.835 0.872 0.511 Overall accuracy 0.855 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 0.9376 0.0025 0.0166 0.0226 0.0004 0.0203 building 0.0146 0.9779 0.0022 0.0016 0.0001 0.0037 low_veg 0.0515 0.0056 0.8120 0.1154 0.0000 0.0156 tree 0.0396 0.0054 0.0608 0.8877 0.0027 0.0039 car 0.0143 0.0114 0.0089 0.0106 0.9252 0.0297 clutter 0.2004 0.1423 0.1692 0.0390 0.0047 0.4445 Precision/Correctness 0.877 0.953 0.858 0.828 0.941 0.737 Recall/Completeness 0.938 0.978 0.812 0.888 0.925 0.444 F1 0.906 0.966 0.834 0.857 0.933 0.554 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 0.9256 0.0093 0.0348 0.0210 0.0014 0.0079 building 0.0142 0.9690 0.0061 0.0053 0.0003 0.0050 low_veg 0.0519 0.0038 0.8080 0.1190 0.0001 0.0172 tree 0.0377 0.0038 0.0823 0.8707 0.0024 0.0030 car 0.1214 0.0011 0.0031 0.0128 0.8331 0.0285 clutter 0.2568 0.1083 0.1361 0.0178 0.0241 0.4570 Precision/Correctness 0.910 0.951 0.770 0.781 0.894 0.801 Recall/Completeness 0.926 0.969 0.808 0.871 0.833 0.457 F1 0.918 0.960 0.788 0.824 0.862 0.582 Overall accuracy 0.878 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 0.9399 0.0059 0.0305 0.0161 0.0005 0.0070 building 0.0100 0.9774 0.0041 0.0040 0.0002 0.0043 low_veg 0.0351 0.0008 0.8358 0.1122 0.0000 0.0161 tree 0.0277 0.0024 0.0723 0.8929 0.0020 0.0026 car 0.0230 0.0010 0.0003 0.0096 0.9335 0.0326 clutter 0.2531 0.0776 0.1294 0.0154 0.0264 0.4981 Precision/Correctness 0.932 0.970 0.792 0.811 0.900 0.821 Recall/Completeness 0.940 0.977 0.836 0.893 0.933 0.498 F1 0.936 0.973 0.813 0.850 0.916 0.620 Overall accuracy 0.901 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 0.9325 0.0240 0.0194 0.0161 0.0014 0.0066 building 0.0328 0.9410 0.0150 0.0046 0.0004 0.0063 low_veg 0.0228 0.0036 0.9377 0.0345 0.0000 0.0014 tree 0.0814 0.0028 0.0978 0.8148 0.0013 0.0020 car 0.0795 0.0056 0.0001 0.0076 0.8725 0.0347 clutter 0.4826 0.0314 0.1553 0.0322 0.0105 0.2880 Precision/Correctness 0.912 0.954 0.896 0.858 0.916 0.541 Recall/Completeness 0.932 0.941 0.938 0.815 0.873 0.288 F1 0.922 0.947 0.916 0.836 0.894 0.376 Overall accuracy 0.908 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 0.9465 0.0219 0.0152 0.0106 0.0004 0.0053 building 0.0286 0.9505 0.0110 0.0041 0.0004 0.0054 low_veg 0.0140 0.0018 0.9544 0.0289 0.0000 0.0009 tree 0.0635 0.0016 0.0857 0.8465 0.0010 0.0017 car 0.0125 0.0064 0.0000 0.0053 0.9398 0.0360 clutter 0.5081 0.0280 0.1369 0.0310 0.0095 0.2865 Precision/Correctness 0.929 0.961 0.915 0.887 0.947 0.561 Recall/Completeness 0.947 0.951 0.954 0.847 0.940 0.287 F1 0.938 0.956 0.934 0.866 0.943 0.379 Overall accuracy 0.926 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 0.9589 0.0044 0.0162 0.0111 0.0034 0.0060 building 0.0370 0.9496 0.0046 0.0016 0.0002 0.0071 low_veg 0.0664 0.0050 0.8402 0.0821 0.0003 0.0059 tree 0.0501 0.0090 0.0773 0.8571 0.0055 0.0010 car 0.0720 0.0032 0.0016 0.0097 0.9040 0.0093 clutter 0.3274 0.1474 0.1977 0.0198 0.0177 0.2900 Precision/Correctness 0.913 0.977 0.868 0.817 0.923 0.537 Recall/Completeness 0.959 0.950 0.840 0.857 0.904 0.290 F1 0.935 0.963 0.854 0.837 0.914 0.377 Overall accuracy 0.911 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 0.9722 0.0017 0.0122 0.0081 0.0002 0.0055 building 0.0320 0.9582 0.0021 0.0011 0.0001 0.0064 low_veg 0.0481 0.0015 0.8701 0.0760 0.0000 0.0043 tree 0.0384 0.0077 0.0666 0.8820 0.0044 0.0008 car 0.0091 0.0035 0.0002 0.0060 0.9715 0.0097 clutter 0.3146 0.1642 0.1939 0.0145 0.0166 0.2962 Precision/Correctness 0.933 0.983 0.891 0.845 0.965 0.544 Recall/Completeness 0.972 0.958 0.870 0.882 0.972 0.296 F1 0.952 0.971 0.880 0.863 0.968 0.384 Overall accuracy 0.930 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 0.9379 0.0051 0.0254 0.0110 0.0036 0.0170 building 0.0165 0.9669 0.0074 0.0027 0.0001 0.0064 low_veg 0.0692 0.0025 0.8187 0.0649 0.0008 0.0438 tree 0.0508 0.0124 0.1426 0.7845 0.0048 0.0048 car 0.0854 0.0016 0.0037 0.0112 0.8842 0.0139 clutter 0.1233 0.0181 0.1040 0.0066 0.0148 0.7332 Precision/Correctness 0.917 0.977 0.753 0.889 0.896 0.738 Recall/Completeness 0.938 0.967 0.819 0.785 0.884 0.733 F1 0.927 0.972 0.785 0.834 0.890 0.735 Overall accuracy 0.889 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 0.9541 0.0026 0.0201 0.0075 0.0002 0.0154 building 0.0119 0.9766 0.0042 0.0019 0.0001 0.0053 low_veg 0.0501 0.0007 0.8497 0.0559 0.0001 0.0435 tree 0.0408 0.0116 0.1283 0.8111 0.0040 0.0041 car 0.0174 0.0017 0.0005 0.0076 0.9583 0.0145 clutter 0.1131 0.0174 0.0946 0.0043 0.0139 0.7566 Precision/Correctness 0.939 0.983 0.777 0.916 0.938 0.760 Recall/Completeness 0.954 0.977 0.850 0.811 0.958 0.757 F1 0.947 0.980 0.811 0.860 0.948 0.758 Overall accuracy 0.911 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 0.9119 0.0322 0.0108 0.0276 0.0015 0.0158 building 0.0172 0.9408 0.0023 0.0048 0.0002 0.0348 low_veg 0.0500 0.0226 0.7373 0.1769 0.0002 0.0129 tree 0.0197 0.0029 0.0945 0.8813 0.0011 0.0005 car 0.0628 0.0016 0.0006 0.0233 0.8886 0.0232 clutter 0.1623 0.0101 0.1174 0.0042 0.0097 0.6964 Precision/Correctness 0.946 0.916 0.758 0.673 0.872 0.784 Recall/Completeness 0.912 0.941 0.737 0.881 0.889 0.696 F1 0.929 0.928 0.747 0.763 0.880 0.738 Overall accuracy 0.879 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 0.9256 0.0292 0.0089 0.0207 0.0004 0.0152 building 0.0143 0.9465 0.0012 0.0037 0.0002 0.0340 low_veg 0.0365 0.0202 0.7617 0.1692 0.0001 0.0125 tree 0.0107 0.0015 0.0872 0.8995 0.0006 0.0004 car 0.0075 0.0012 0.0001 0.0159 0.9494 0.0258 clutter 0.1533 0.0088 0.1173 0.0039 0.0091 0.7076 Precision/Correctness 0.955 0.925 0.770 0.702 0.900 0.794 Recall/Completeness 0.926 0.946 0.762 0.900 0.949 0.708 F1 0.940 0.936 0.766 0.789 0.924 0.748 Overall accuracy 0.893 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image