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 0.907 0.017 0.039 0.017 0.007 0.013 building 0.016 0.965 0.009 0.003 0.000 0.007 low_veg 0.046 0.011 0.874 0.058 0.001 0.010 tree 0.038 0.006 0.134 0.816 0.003 0.003 car 0.113 0.005 0.006 0.014 0.848 0.014 clutter 0.281 0.106 0.163 0.022 0.008 0.420 Precision/Correctness 0.891 0.944 0.804 0.877 0.836 0.687 Recall/Completeness 0.907 0.965 0.874 0.816 0.848 0.420 F1 0.899 0.955 0.837 0.845 0.842 0.521 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 0.930 0.012 0.031 0.012 0.003 0.012 building 0.012 0.974 0.006 0.002 0.000 0.006 low_veg 0.033 0.006 0.902 0.049 0.000 0.009 tree 0.028 0.004 0.120 0.843 0.002 0.002 car 0.044 0.005 0.003 0.009 0.923 0.016 clutter 0.277 0.090 0.145 0.018 0.008 0.462 Precision/Correctness 0.916 0.962 0.831 0.904 0.904 0.711 Recall/Completeness 0.930 0.974 0.902 0.843 0.923 0.462 F1 0.923 0.968 0.865 0.873 0.913 0.560 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 0.8858 0.0213 0.0478 0.0126 0.0047 0.0277 building 0.0117 0.9661 0.0111 0.0063 0.0000 0.0048 low_veg 0.0313 0.0110 0.9049 0.0396 0.0002 0.0129 tree 0.0388 0.0099 0.1426 0.8025 0.0016 0.0046 car 0.1172 0.0021 0.0056 0.0104 0.8205 0.0441 clutter 0.1541 0.1875 0.1655 0.0242 0.0022 0.4665 Precision/Correctness 0.897 0.925 0.845 0.917 0.854 0.405 Recall/Completeness 0.886 0.966 0.905 0.803 0.820 0.467 F1 0.891 0.945 0.874 0.856 0.837 0.434 Overall accuracy 0.878 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 0.9108 0.0154 0.0384 0.0082 0.0012 0.0260 building 0.0078 0.9768 0.0063 0.0051 0.0000 0.0040 low_veg 0.0216 0.0062 0.9314 0.0291 0.0001 0.0116 tree 0.0293 0.0080 0.1218 0.8361 0.0011 0.0037 car 0.0496 0.0014 0.0014 0.0053 0.8889 0.0533 clutter 0.1093 0.1886 0.1190 0.0163 0.0019 0.5650 Precision/Correctness 0.930 0.948 0.874 0.942 0.930 0.401 Recall/Completeness 0.911 0.977 0.931 0.836 0.889 0.565 F1 0.920 0.962 0.902 0.886 0.909 0.469 Overall accuracy 0.907 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 0.8891 0.0072 0.0844 0.0112 0.0044 0.0037 building 0.0068 0.9722 0.0116 0.0060 0.0001 0.0033 low_veg 0.0209 0.0034 0.9356 0.0328 0.0002 0.0071 tree 0.0181 0.0029 0.1998 0.7773 0.0009 0.0009 car 0.1046 0.0076 0.0041 0.0084 0.8750 0.0002 clutter 0.1133 0.2342 0.2340 0.0719 0.0000 0.3466 Precision/Correctness 0.892 0.938 0.819 0.945 0.828 0.236 Recall/Completeness 0.889 0.972 0.936 0.777 0.875 0.347 F1 0.891 0.955 0.873 0.853 0.851 0.281 Overall accuracy 0.872 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 0.9206 0.0036 0.0648 0.0067 0.0013 0.0030 building 0.0037 0.9849 0.0048 0.0038 0.0000 0.0027 low_veg 0.0147 0.0014 0.9508 0.0259 0.0001 0.0071 tree 0.0139 0.0022 0.1835 0.7990 0.0007 0.0007 car 0.0414 0.0086 0.0006 0.0048 0.9444 0.0002 clutter 0.0806 0.2600 0.1692 0.0831 0.0000 0.4071 Precision/Correctness 0.921 0.963 0.837 0.959 0.907 0.214 Recall/Completeness 0.921 0.985 0.951 0.799 0.944 0.407 F1 0.921 0.974 0.891 0.872 0.925 0.280 Overall accuracy 0.893 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 0.8887 0.0211 0.0483 0.0190 0.0086 0.0143 building 0.0113 0.9736 0.0070 0.0032 0.0000 0.0049 low_veg 0.0522 0.0218 0.8627 0.0540 0.0004 0.0088 tree 0.0312 0.0078 0.1151 0.8409 0.0025 0.0024 car 0.1083 0.0073 0.0039 0.0138 0.8595 0.0072 clutter 0.2830 0.1504 0.2824 0.0238 0.0028 0.2577 Precision/Correctness 0.879 0.938 0.808 0.912 0.837 0.449 Recall/Completeness 0.889 0.974 0.863 0.841 0.859 0.258 F1 0.884 0.956 0.835 0.875 0.848 0.327 Overall accuracy 0.877 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 0.9203 0.0142 0.0374 0.0122 0.0025 0.0133 building 0.0080 0.9828 0.0027 0.0022 0.0000 0.0043 low_veg 0.0385 0.0122 0.8980 0.0438 0.0001 0.0074 tree 0.0229 0.0062 0.0973 0.8696 0.0018 0.0021 car 0.0387 0.0082 0.0021 0.0090 0.9344 0.0076 clutter 0.2866 0.1451 0.2405 0.0154 0.0034 0.3091 Precision/Correctness 0.913 0.960 0.844 0.937 0.918 0.459 Recall/Completeness 0.920 0.983 0.898 0.870 0.934 0.309 F1 0.917 0.971 0.870 0.902 0.926 0.370 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 0.8773 0.0166 0.0656 0.0202 0.0066 0.0136 building 0.0118 0.9755 0.0071 0.0030 0.0000 0.0026 low_veg 0.0338 0.0110 0.9020 0.0450 0.0001 0.0081 tree 0.0244 0.0043 0.1910 0.7753 0.0014 0.0036 car 0.1042 0.0124 0.0021 0.0299 0.8400 0.0113 clutter 0.2114 0.1678 0.2054 0.0255 0.0044 0.3855 Precision/Correctness 0.878 0.944 0.775 0.917 0.840 0.636 Recall/Completeness 0.877 0.975 0.902 0.775 0.840 0.386 F1 0.878 0.959 0.834 0.840 0.840 0.480 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 0.9100 0.0099 0.0528 0.0131 0.0019 0.0123 building 0.0091 0.9845 0.0026 0.0019 0.0000 0.0018 low_veg 0.0215 0.0054 0.9316 0.0344 0.0000 0.0070 tree 0.0189 0.0032 0.1755 0.7983 0.0011 0.0031 car 0.0356 0.0133 0.0006 0.0208 0.9161 0.0136 clutter 0.2013 0.1618 0.1679 0.0184 0.0044 0.4463 Precision/Correctness 0.911 0.963 0.801 0.942 0.916 0.664 Recall/Completeness 0.910 0.985 0.932 0.798 0.916 0.446 F1 0.911 0.974 0.861 0.864 0.916 0.534 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_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9165 0.0112 0.0339 0.0254 0.0084 0.0045 building 0.0120 0.9735 0.0074 0.0034 0.0001 0.0036 low_veg 0.0534 0.0131 0.8197 0.1080 0.0009 0.0048 tree 0.0324 0.0036 0.0494 0.9116 0.0021 0.0009 car 0.1059 0.0011 0.0103 0.0259 0.8493 0.0074 clutter 0.4825 0.1202 0.1811 0.0674 0.0116 0.1372 Precision/Correctness 0.845 0.959 0.790 0.798 0.832 0.724 Recall/Completeness 0.917 0.974 0.820 0.912 0.849 0.137 F1 0.879 0.966 0.805 0.851 0.840 0.231 Overall accuracy 0.867 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 0.9416 0.0066 0.0270 0.0184 0.0030 0.0035 building 0.0069 0.9845 0.0032 0.0022 0.0000 0.0031 low_veg 0.0375 0.0054 0.8571 0.0963 0.0002 0.0035 tree 0.0207 0.0020 0.0371 0.9383 0.0014 0.0005 car 0.0390 0.0005 0.0049 0.0175 0.9306 0.0076 clutter 0.5050 0.1026 0.1703 0.0581 0.0131 0.1509 Precision/Correctness 0.873 0.974 0.829 0.835 0.897 0.764 Recall/Completeness 0.942 0.985 0.857 0.938 0.931 0.151 F1 0.906 0.979 0.843 0.884 0.913 0.252 Overall accuracy 0.896 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 0.8602 0.0199 0.0762 0.0152 0.0109 0.0176 building 0.0083 0.9777 0.0050 0.0044 0.0000 0.0045 low_veg 0.0862 0.0166 0.8122 0.0628 0.0012 0.0210 tree 0.0343 0.0060 0.1300 0.8211 0.0029 0.0057 car 0.1051 0.0034 0.0053 0.0203 0.8443 0.0216 clutter 0.4074 0.0454 0.1616 0.0125 0.0009 0.3722 Precision/Correctness 0.756 0.931 0.754 0.881 0.840 0.757 Recall/Completeness 0.860 0.978 0.812 0.821 0.844 0.372 F1 0.805 0.954 0.782 0.850 0.842 0.499 Overall accuracy 0.817 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 0.8846 0.0147 0.0701 0.0099 0.0039 0.0168 building 0.0055 0.9853 0.0022 0.0032 0.0000 0.0037 low_veg 0.0719 0.0106 0.8437 0.0537 0.0004 0.0197 tree 0.0250 0.0044 0.1134 0.8500 0.0022 0.0049 car 0.0471 0.0031 0.0022 0.0145 0.9097 0.0234 clutter 0.4171 0.0413 0.1469 0.0095 0.0006 0.3846 Precision/Correctness 0.775 0.949 0.776 0.908 0.917 0.781 Recall/Completeness 0.885 0.985 0.844 0.850 0.910 0.385 F1 0.826 0.967 0.809 0.878 0.913 0.515 Overall accuracy 0.840 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 0.9013 0.0205 0.0558 0.0060 0.0074 0.0089 building 0.0048 0.9840 0.0067 0.0022 0.0000 0.0023 low_veg 0.0653 0.0224 0.8698 0.0327 0.0017 0.0080 tree 0.0744 0.0100 0.1440 0.7595 0.0075 0.0045 car 0.1179 0.0061 0.0083 0.0051 0.8443 0.0183 clutter 0.3412 0.2288 0.1634 0.0134 0.0020 0.2512 Precision/Correctness 0.899 0.942 0.762 0.919 0.840 0.544 Recall/Completeness 0.901 0.984 0.870 0.759 0.844 0.251 F1 0.900 0.963 0.812 0.832 0.842 0.344 Overall accuracy 0.883 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 0.9225 0.0139 0.0500 0.0036 0.0025 0.0075 building 0.0024 0.9906 0.0036 0.0016 0.0000 0.0018 low_veg 0.0415 0.0130 0.9134 0.0257 0.0006 0.0058 tree 0.0572 0.0078 0.1242 0.8014 0.0057 0.0036 car 0.0450 0.0061 0.0031 0.0022 0.9210 0.0225 clutter 0.3485 0.2320 0.1349 0.0098 0.0016 0.2732 Precision/Correctness 0.930 0.960 0.793 0.944 0.910 0.568 Recall/Completeness 0.923 0.991 0.913 0.801 0.921 0.273 F1 0.926 0.975 0.849 0.867 0.916 0.369 Overall accuracy 0.912 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 0.9206 0.0160 0.0286 0.0215 0.0058 0.0074 building 0.0088 0.9780 0.0064 0.0033 0.0000 0.0033 low_veg 0.0607 0.0120 0.8526 0.0650 0.0001 0.0096 tree 0.0491 0.0053 0.1131 0.8277 0.0026 0.0022 car 0.1271 0.0011 0.0018 0.0092 0.8544 0.0063 clutter 0.2862 0.2466 0.1533 0.0147 0.0459 0.2533 Precision/Correctness 0.904 0.945 0.843 0.828 0.854 0.588 Recall/Completeness 0.921 0.978 0.853 0.828 0.854 0.253 F1 0.912 0.961 0.848 0.828 0.854 0.354 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: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9424 0.0101 0.0241 0.0161 0.0016 0.0057 building 0.0049 0.9849 0.0047 0.0027 0.0000 0.0029 low_veg 0.0529 0.0067 0.8766 0.0561 0.0000 0.0077 tree 0.0337 0.0033 0.0985 0.8604 0.0022 0.0018 car 0.0462 0.0009 0.0009 0.0068 0.9382 0.0071 clutter 0.2552 0.2352 0.1458 0.0120 0.0588 0.2931 Precision/Correctness 0.932 0.964 0.870 0.858 0.898 0.615 Recall/Completeness 0.942 0.985 0.877 0.860 0.938 0.293 F1 0.937 0.974 0.873 0.859 0.918 0.397 Overall accuracy 0.916 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 0.9223 0.0077 0.0272 0.0248 0.0083 0.0098 building 0.0168 0.9734 0.0047 0.0022 0.0001 0.0029 low_veg 0.0571 0.0090 0.8362 0.0855 0.0008 0.0113 tree 0.0498 0.0064 0.0872 0.8473 0.0044 0.0048 car 0.1141 0.0060 0.0161 0.0136 0.8304 0.0197 clutter 0.1902 0.1803 0.1794 0.0402 0.0050 0.4049 Precision/Correctness 0.858 0.929 0.822 0.841 0.820 0.813 Recall/Completeness 0.922 0.973 0.836 0.847 0.830 0.405 F1 0.889 0.951 0.829 0.844 0.825 0.541 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_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9466 0.0042 0.0189 0.0189 0.0033 0.0081 building 0.0146 0.9788 0.0022 0.0017 0.0001 0.0026 low_veg 0.0413 0.0051 0.8637 0.0801 0.0003 0.0096 tree 0.0390 0.0050 0.0779 0.8705 0.0035 0.0041 car 0.0452 0.0069 0.0100 0.0089 0.9065 0.0225 clutter 0.1772 0.1534 0.1680 0.0385 0.0048 0.4582 Precision/Correctness 0.888 0.950 0.850 0.863 0.884 0.841 Recall/Completeness 0.947 0.979 0.864 0.870 0.907 0.458 F1 0.916 0.964 0.857 0.867 0.895 0.593 Overall accuracy 0.891 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 0.9023 0.0115 0.0447 0.0171 0.0052 0.0191 building 0.0145 0.9636 0.0091 0.0052 0.0002 0.0074 low_veg 0.0517 0.0074 0.8194 0.1087 0.0002 0.0127 tree 0.0379 0.0040 0.0834 0.8699 0.0026 0.0022 car 0.1405 0.0061 0.0085 0.0104 0.8205 0.0140 clutter 0.2428 0.1116 0.1304 0.0149 0.0028 0.4975 Precision/Correctness 0.909 0.944 0.750 0.802 0.891 0.746 Recall/Completeness 0.902 0.964 0.819 0.870 0.821 0.497 F1 0.906 0.954 0.783 0.834 0.854 0.597 Overall accuracy 0.871 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 0.9216 0.0069 0.0386 0.0129 0.0016 0.0183 building 0.0097 0.9728 0.0066 0.0040 0.0001 0.0068 low_veg 0.0325 0.0032 0.8492 0.1027 0.0000 0.0123 tree 0.0266 0.0018 0.0730 0.8947 0.0019 0.0019 car 0.0580 0.0069 0.0040 0.0088 0.9058 0.0164 clutter 0.2338 0.0810 0.1151 0.0125 0.0027 0.5548 Precision/Correctness 0.933 0.966 0.778 0.830 0.946 0.761 Recall/Completeness 0.922 0.973 0.849 0.895 0.906 0.555 F1 0.927 0.969 0.812 0.861 0.926 0.642 Overall accuracy 0.897 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_6_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9206 0.0235 0.0233 0.0157 0.0041 0.0128 building 0.0288 0.9471 0.0180 0.0014 0.0002 0.0046 low_veg 0.0219 0.0063 0.9373 0.0324 0.0000 0.0020 tree 0.0754 0.0039 0.0921 0.8254 0.0014 0.0018 car 0.1083 0.0252 0.0001 0.0046 0.8442 0.0178 clutter 0.3824 0.0420 0.2995 0.0155 0.0025 0.2581 Precision/Correctness 0.920 0.950 0.880 0.873 0.851 0.437 Recall/Completeness 0.921 0.947 0.937 0.825 0.844 0.258 F1 0.920 0.949 0.908 0.849 0.847 0.324 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 0.9396 0.0200 0.0171 0.0103 0.0015 0.0115 building 0.0249 0.9575 0.0128 0.0008 0.0001 0.0038 low_veg 0.0122 0.0034 0.9563 0.0269 0.0000 0.0012 tree 0.0560 0.0024 0.0804 0.8585 0.0010 0.0016 car 0.0407 0.0305 0.0000 0.0020 0.9065 0.0204 clutter 0.3807 0.0393 0.2973 0.0137 0.0016 0.2674 Precision/Correctness 0.939 0.961 0.902 0.903 0.923 0.450 Recall/Completeness 0.940 0.957 0.956 0.859 0.906 0.267 F1 0.939 0.959 0.929 0.880 0.915 0.335 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 0.9428 0.0076 0.0247 0.0101 0.0111 0.0037 building 0.0435 0.9344 0.0079 0.0015 0.0005 0.0122 low_veg 0.0554 0.0135 0.8581 0.0656 0.0012 0.0061 tree 0.0488 0.0087 0.0922 0.8440 0.0056 0.0008 car 0.1048 0.0025 0.0038 0.0120 0.8688 0.0082 clutter 0.2575 0.1323 0.2927 0.0200 0.0122 0.2853 Precision/Correctness 0.912 0.969 0.827 0.839 0.833 0.509 Recall/Completeness 0.943 0.934 0.858 0.844 0.869 0.285 F1 0.927 0.951 0.842 0.842 0.851 0.366 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 0.9632 0.0039 0.0189 0.0073 0.0038 0.0030 building 0.0383 0.9440 0.0045 0.0009 0.0002 0.0120 low_veg 0.0367 0.0082 0.8903 0.0598 0.0004 0.0046 tree 0.0372 0.0073 0.0793 0.8713 0.0042 0.0006 car 0.0346 0.0014 0.0014 0.0084 0.9453 0.0088 clutter 0.2316 0.1424 0.2880 0.0160 0.0124 0.3095 Precision/Correctness 0.934 0.979 0.856 0.867 0.911 0.521 Recall/Completeness 0.963 0.944 0.890 0.871 0.945 0.310 F1 0.949 0.961 0.873 0.869 0.928 0.388 Overall accuracy 0.923 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 0.9194 0.0068 0.0334 0.0099 0.0132 0.0173 building 0.0127 0.9525 0.0251 0.0026 0.0000 0.0071 low_veg 0.0451 0.0052 0.8718 0.0468 0.0029 0.0282 tree 0.0418 0.0059 0.1622 0.7818 0.0046 0.0038 car 0.1047 0.0018 0.0102 0.0100 0.8647 0.0086 clutter 0.1315 0.0109 0.1610 0.0095 0.0103 0.6768 Precision/Correctness 0.927 0.978 0.707 0.909 0.796 0.754 Recall/Completeness 0.919 0.953 0.872 0.782 0.865 0.677 F1 0.923 0.965 0.781 0.841 0.829 0.713 Overall accuracy 0.883 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 0.9439 0.0033 0.0246 0.0066 0.0059 0.0158 building 0.0075 0.9627 0.0215 0.0017 0.0000 0.0066 low_veg 0.0269 0.0017 0.9049 0.0382 0.0012 0.0272 tree 0.0311 0.0044 0.1457 0.8119 0.0036 0.0033 car 0.0434 0.0016 0.0032 0.0062 0.9361 0.0095 clutter 0.1229 0.0094 0.1465 0.0071 0.0106 0.7036 Precision/Correctness 0.950 0.988 0.734 0.935 0.866 0.776 Recall/Completeness 0.944 0.963 0.905 0.812 0.936 0.704 F1 0.947 0.975 0.810 0.869 0.900 0.738 Overall accuracy 0.908 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 0.9075 0.0323 0.0096 0.0239 0.0041 0.0226 building 0.0183 0.9471 0.0032 0.0017 0.0002 0.0295 low_veg 0.0643 0.0146 0.7752 0.1370 0.0006 0.0083 tree 0.0309 0.0087 0.0958 0.8628 0.0009 0.0010 car 0.1073 0.0041 0.0025 0.0222 0.8466 0.0173 clutter 0.2150 0.0203 0.0541 0.0017 0.0092 0.6997 Precision/Correctness 0.930 0.914 0.815 0.717 0.793 0.769 Recall/Completeness 0.908 0.947 0.775 0.863 0.847 0.700 F1 0.919 0.930 0.794 0.783 0.819 0.733 Overall accuracy 0.880 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 0.9233 0.0288 0.0069 0.0173 0.0016 0.0221 building 0.0156 0.9525 0.0022 0.0010 0.0002 0.0286 low_veg 0.0477 0.0096 0.8063 0.1289 0.0002 0.0073 tree 0.0195 0.0063 0.0878 0.8853 0.0005 0.0007 car 0.0392 0.0049 0.0007 0.0162 0.9193 0.0196 clutter 0.2069 0.0184 0.0523 0.0012 0.0089 0.7124 Precision/Correctness 0.941 0.926 0.836 0.751 0.853 0.779 Recall/Completeness 0.923 0.952 0.806 0.885 0.919 0.712 F1 0.932 0.939 0.821 0.812 0.885 0.744 Overall accuracy 0.897 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image