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 0.882 0.035 0.043 0.022 0.003 0.016 building 0.042 0.916 0.016 0.009 0.001 0.016 low_veg 0.048 0.012 0.849 0.080 0.000 0.011 tree 0.048 0.007 0.178 0.757 0.003 0.006 car 0.077 0.014 0.004 0.015 0.868 0.022 clutter 0.316 0.133 0.181 0.031 0.013 0.325 Precision/Correctness 0.862 0.912 0.761 0.825 0.882 0.544 Recall/Completeness 0.882 0.916 0.849 0.757 0.868 0.325 F1 0.872 0.914 0.803 0.789 0.875 0.407 Overall accuracy 0.836	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 0.900 0.032 0.036 0.017 0.000 0.015 building 0.039 0.924 0.013 0.008 0.001 0.015 low_veg 0.037 0.008 0.873 0.072 0.000 0.009 tree 0.039 0.006 0.165 0.782 0.002 0.006 car 0.017 0.015 0.002 0.011 0.931 0.024 clutter 0.320 0.120 0.173 0.027 0.013 0.346 Precision/Correctness 0.882 0.927 0.782 0.850 0.919 0.562 Recall/Completeness 0.900 0.924 0.873 0.782 0.931 0.346 F1 0.891 0.926 0.825 0.814 0.925 0.428 Overall accuracy 0.857
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8696 0.0264 0.0633 0.0115 0.0025 0.0268 building 0.0589 0.8561 0.0455 0.0099 0.0002 0.0294 low_veg 0.0325 0.0094 0.8943 0.0473 0.0001 0.0163 tree 0.0471 0.0106 0.1995 0.7262 0.0017 0.0151 car 0.0975 0.0147 0.0052 0.0069 0.8454 0.0303 clutter 0.1429 0.2248 0.1444 0.0340 0.0288 0.4251 Precision/Correctness 0.858 0.906 0.788 0.895 0.858 0.282 Recall/Completeness 0.870 0.856 0.894 0.726 0.845 0.425 F1 0.864 0.881 0.838 0.802 0.852 0.339 Overall accuracy 0.832 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 0.8920 0.0221 0.0534 0.0079 0.0003 0.0242 building 0.0554 0.8663 0.0419 0.0090 0.0002 0.0273 low_veg 0.0255 0.0061 0.9162 0.0382 0.0001 0.0139 tree 0.0385 0.0092 0.1799 0.7568 0.0012 0.0145 car 0.0418 0.0169 0.0020 0.0030 0.9029 0.0333 clutter 0.1147 0.2313 0.1124 0.0294 0.0344 0.4778 Precision/Correctness 0.882 0.928 0.811 0.919 0.904 0.263 Recall/Completeness 0.892 0.866 0.916 0.757 0.903 0.478 F1 0.887 0.896 0.860 0.830 0.904 0.339 Overall accuracy 0.857 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 0.8694 0.0189 0.0901 0.0091 0.0015 0.0110 building 0.0482 0.8993 0.0179 0.0076 0.0011 0.0259 low_veg 0.0166 0.0035 0.9402 0.0312 0.0001 0.0085 tree 0.0314 0.0041 0.2909 0.6652 0.0010 0.0074 car 0.0831 0.0285 0.0018 0.0071 0.8760 0.0035 clutter 0.1012 0.3103 0.2253 0.0449 0.0352 0.2831 Precision/Correctness 0.858 0.897 0.765 0.940 0.864 0.095 Recall/Completeness 0.869 0.899 0.940 0.665 0.876 0.283 F1 0.864 0.898 0.844 0.779 0.870 0.143 Overall accuracy 0.826 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_2_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8975 0.0165 0.0703 0.0056 0.0002 0.0099 building 0.0470 0.9099 0.0119 0.0061 0.0010 0.0242 low_veg 0.0120 0.0018 0.9525 0.0258 0.0000 0.0080 tree 0.0275 0.0035 0.2763 0.6848 0.0008 0.0072 car 0.0236 0.0317 0.0000 0.0035 0.9377 0.0035 clutter 0.0760 0.3387 0.2028 0.0476 0.0345 0.3003 Precision/Correctness 0.879 0.921 0.780 0.953 0.903 0.078 Recall/Completeness 0.897 0.910 0.952 0.685 0.938 0.300 F1 0.888 0.915 0.858 0.797 0.920 0.124 Overall accuracy 0.843 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_3_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8746 0.0218 0.0671 0.0192 0.0035 0.0138 building 0.0354 0.9336 0.0129 0.0039 0.0002 0.0140 low_veg 0.0509 0.0198 0.8547 0.0617 0.0002 0.0127 tree 0.0387 0.0094 0.1509 0.7918 0.0028 0.0063 car 0.0761 0.0119 0.0032 0.0161 0.8782 0.0144 clutter 0.2948 0.1426 0.2802 0.0355 0.0140 0.2329 Precision/Correctness 0.853 0.936 0.762 0.896 0.884 0.324 Recall/Completeness 0.875 0.934 0.855 0.792 0.878 0.233 F1 0.864 0.935 0.806 0.841 0.881 0.271 Overall accuracy 0.850 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 0.9021 0.0161 0.0559 0.0136 0.0003 0.0120 building 0.0324 0.9426 0.0096 0.0029 0.0001 0.0124 low_veg 0.0408 0.0116 0.8852 0.0524 0.0001 0.0098 tree 0.0310 0.0078 0.1349 0.8184 0.0022 0.0056 car 0.0148 0.0134 0.0009 0.0114 0.9436 0.0159 clutter 0.3080 0.1381 0.2649 0.0285 0.0163 0.2441 Precision/Correctness 0.880 0.956 0.789 0.921 0.932 0.311 Recall/Completeness 0.902 0.943 0.885 0.818 0.944 0.244 F1 0.891 0.949 0.834 0.867 0.938 0.273 Overall accuracy 0.878 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 0.8473 0.0350 0.0774 0.0187 0.0038 0.0178 building 0.0372 0.9331 0.0098 0.0032 0.0002 0.0166 low_veg 0.0310 0.0113 0.8973 0.0527 0.0001 0.0075 tree 0.0353 0.0079 0.2685 0.6782 0.0019 0.0083 car 0.0704 0.0145 0.0035 0.0282 0.8543 0.0291 clutter 0.2113 0.1888 0.1784 0.0251 0.0091 0.3872 Precision/Correctness 0.844 0.918 0.723 0.897 0.863 0.496 Recall/Completeness 0.847 0.933 0.897 0.678 0.854 0.387 F1 0.846 0.925 0.801 0.773 0.859 0.435 Overall accuracy 0.822 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 0.8751 0.0305 0.0649 0.0130 0.0007 0.0158 building 0.0351 0.9408 0.0064 0.0023 0.0002 0.0152 low_veg 0.0214 0.0065 0.9230 0.0434 0.0001 0.0056 tree 0.0301 0.0071 0.2555 0.6984 0.0016 0.0074 car 0.0143 0.0162 0.0014 0.0213 0.9155 0.0312 clutter 0.2054 0.1834 0.1490 0.0181 0.0104 0.4338 Precision/Correctness 0.870 0.936 0.741 0.922 0.908 0.513 Recall/Completeness 0.875 0.941 0.923 0.698 0.915 0.434 F1 0.872 0.939 0.822 0.795 0.911 0.470 Overall accuracy 0.846 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 0.9117 0.0163 0.0375 0.0265 0.0021 0.0059 building 0.0276 0.9504 0.0074 0.0065 0.0003 0.0078 low_veg 0.0551 0.0092 0.8098 0.1190 0.0003 0.0065 tree 0.0473 0.0040 0.0737 0.8709 0.0020 0.0021 car 0.0729 0.0041 0.0073 0.0247 0.8581 0.0328 clutter 0.5074 0.1331 0.1855 0.0616 0.0114 0.1009 Precision/Correctness 0.824 0.952 0.766 0.776 0.912 0.520 Recall/Completeness 0.912 0.950 0.810 0.871 0.858 0.101 F1 0.865 0.951 0.787 0.821 0.884 0.169 Overall accuracy 0.848 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 0.9307 0.0133 0.0310 0.0201 0.0003 0.0046 building 0.0227 0.9616 0.0036 0.0053 0.0002 0.0066 low_veg 0.0433 0.0050 0.8370 0.1095 0.0002 0.0050 tree 0.0357 0.0025 0.0612 0.8976 0.0014 0.0016 car 0.0103 0.0043 0.0027 0.0176 0.9306 0.0344 clutter 0.5323 0.1260 0.1747 0.0547 0.0119 0.1003 Precision/Correctness 0.847 0.963 0.799 0.810 0.937 0.540 Recall/Completeness 0.931 0.962 0.837 0.898 0.931 0.100 F1 0.887 0.962 0.818 0.852 0.934 0.169 Overall accuracy 0.874 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 0.8597 0.0326 0.0661 0.0178 0.0038 0.0201 building 0.0217 0.9543 0.0073 0.0064 0.0003 0.0099 low_veg 0.0850 0.0190 0.7689 0.0994 0.0007 0.0270 tree 0.0491 0.0081 0.1525 0.7791 0.0032 0.0079 car 0.0720 0.0204 0.0038 0.0194 0.8648 0.0196 clutter 0.4615 0.0614 0.1461 0.0149 0.0024 0.3137 Precision/Correctness 0.731 0.903 0.742 0.827 0.906 0.670 Recall/Completeness 0.860 0.954 0.769 0.779 0.865 0.314 F1 0.790 0.928 0.755 0.803 0.885 0.427 Overall accuracy 0.789 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 0.8777 0.0280 0.0619 0.0132 0.0004 0.0187 building 0.0189 0.9617 0.0053 0.0053 0.0003 0.0086 low_veg 0.0728 0.0137 0.7953 0.0919 0.0003 0.0260 tree 0.0402 0.0064 0.1380 0.8058 0.0025 0.0070 car 0.0240 0.0217 0.0020 0.0149 0.9173 0.0201 clutter 0.4744 0.0575 0.1350 0.0112 0.0017 0.3201 Precision/Correctness 0.744 0.921 0.759 0.853 0.951 0.696 Recall/Completeness 0.878 0.962 0.795 0.806 0.917 0.320 F1 0.805 0.941 0.777 0.829 0.934 0.439 Overall accuracy 0.808 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 0.8940 0.0225 0.0559 0.0116 0.0034 0.0127 building 0.0827 0.8510 0.0188 0.0217 0.0004 0.0255 low_veg 0.0621 0.0329 0.8319 0.0609 0.0007 0.0115 tree 0.0844 0.0092 0.1701 0.7177 0.0087 0.0100 car 0.0866 0.0174 0.0048 0.0082 0.8560 0.0270 clutter 0.3264 0.2174 0.1662 0.0244 0.0122 0.2533 Precision/Correctness 0.840 0.927 0.725 0.810 0.877 0.313 Recall/Completeness 0.894 0.851 0.832 0.718 0.856 0.253 F1 0.866 0.887 0.775 0.761 0.867 0.280 Overall accuracy 0.828 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 0.9107 0.0183 0.0503 0.0086 0.0009 0.0111 building 0.0800 0.8580 0.0165 0.0211 0.0003 0.0241 low_veg 0.0428 0.0277 0.8661 0.0546 0.0002 0.0086 tree 0.0694 0.0078 0.1524 0.7538 0.0073 0.0093 car 0.0224 0.0205 0.0009 0.0045 0.9232 0.0284 clutter 0.3388 0.2256 0.1604 0.0195 0.0114 0.2443 Precision/Correctness 0.862 0.941 0.747 0.835 0.914 0.284 Recall/Completeness 0.911 0.858 0.866 0.754 0.923 0.244 F1 0.886 0.898 0.802 0.792 0.919 0.263 Overall accuracy 0.851 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 0.9050 0.0252 0.0317 0.0276 0.0023 0.0083 building 0.0169 0.9632 0.0054 0.0039 0.0003 0.0104 low_veg 0.0707 0.0149 0.8029 0.1057 0.0001 0.0056 tree 0.0610 0.0067 0.1224 0.8044 0.0027 0.0029 car 0.0927 0.0092 0.0007 0.0099 0.8641 0.0234 clutter 0.2528 0.2861 0.1587 0.0292 0.0640 0.2092 Precision/Correctness 0.892 0.927 0.826 0.758 0.872 0.465 Recall/Completeness 0.905 0.963 0.803 0.804 0.864 0.209 F1 0.899 0.945 0.814 0.781 0.868 0.289 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_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9240 0.0195 0.0271 0.0219 0.0002 0.0072 building 0.0133 0.9699 0.0038 0.0031 0.0002 0.0097 low_veg 0.0638 0.0105 0.8247 0.0970 0.0000 0.0040 tree 0.0462 0.0049 0.1112 0.8329 0.0023 0.0025 car 0.0200 0.0098 0.0000 0.0073 0.9372 0.0257 clutter 0.2271 0.2829 0.1570 0.0278 0.0783 0.2268 Precision/Correctness 0.917 0.946 0.851 0.786 0.894 0.451 Recall/Completeness 0.924 0.970 0.825 0.833 0.937 0.227 F1 0.920 0.958 0.838 0.809 0.915 0.302 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8914 0.0266 0.0299 0.0296 0.0022 0.0203 building 0.0255 0.9547 0.0066 0.0030 0.0007 0.0095 low_veg 0.0777 0.0147 0.7792 0.1193 0.0003 0.0088 tree 0.0612 0.0068 0.1199 0.8035 0.0040 0.0045 car 0.0896 0.0276 0.0092 0.0165 0.8495 0.0076 clutter 0.2233 0.2231 0.1711 0.0573 0.0083 0.3170 Precision/Correctness 0.823 0.894 0.787 0.788 0.885 0.690 Recall/Completeness 0.891 0.955 0.779 0.804 0.849 0.317 F1 0.856 0.923 0.783 0.795 0.867 0.434 Overall accuracy 0.825 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 0.9101 0.0234 0.0233 0.0238 0.0004 0.0190 building 0.0234 0.9605 0.0042 0.0025 0.0007 0.0086 low_veg 0.0637 0.0110 0.8027 0.1159 0.0000 0.0066 tree 0.0505 0.0057 0.1121 0.8248 0.0032 0.0037 car 0.0180 0.0288 0.0092 0.0114 0.9253 0.0072 clutter 0.2174 0.2015 0.1676 0.0553 0.0086 0.3495 Precision/Correctness 0.849 0.915 0.809 0.810 0.911 0.714 Recall/Completeness 0.910 0.961 0.803 0.825 0.925 0.349 F1 0.878 0.937 0.806 0.817 0.918 0.469 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: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8314 0.0604 0.0507 0.0284 0.0022 0.0270 building 0.0509 0.8649 0.0211 0.0407 0.0008 0.0216 low_veg 0.0513 0.0120 0.7579 0.1665 0.0002 0.0121 tree 0.0443 0.0040 0.0981 0.8477 0.0031 0.0029 car 0.1035 0.0048 0.0048 0.0170 0.8353 0.0347 clutter 0.3625 0.1265 0.1626 0.0423 0.0134 0.2927 Precision/Correctness 0.862 0.858 0.691 0.669 0.902 0.518 Recall/Completeness 0.831 0.865 0.758 0.848 0.835 0.293 F1 0.846 0.862 0.723 0.748 0.867 0.374 Overall accuracy 0.793 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 0.8450 0.0586 0.0459 0.0234 0.0008 0.0263 building 0.0474 0.8736 0.0190 0.0396 0.0006 0.0197 low_veg 0.0357 0.0083 0.7817 0.1634 0.0001 0.0108 tree 0.0345 0.0025 0.0885 0.8694 0.0026 0.0024 car 0.0234 0.0052 0.0012 0.0146 0.9167 0.0387 clutter 0.3779 0.0965 0.1581 0.0414 0.0139 0.3122 Precision/Correctness 0.880 0.877 0.710 0.692 0.920 0.528 Recall/Completeness 0.845 0.874 0.782 0.869 0.917 0.312 F1 0.862 0.875 0.744 0.770 0.918 0.392 Overall accuracy 0.813 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 0.9124 0.0260 0.0235 0.0206 0.0015 0.0160 building 0.0526 0.9083 0.0215 0.0026 0.0011 0.0139 low_veg 0.0265 0.0059 0.9179 0.0478 0.0000 0.0018 tree 0.0912 0.0047 0.1238 0.7764 0.0021 0.0018 car 0.0729 0.0291 0.0004 0.0074 0.8355 0.0546 clutter 0.4107 0.0624 0.2821 0.0352 0.0067 0.2029 Precision/Correctness 0.898 0.942 0.863 0.818 0.895 0.280 Recall/Completeness 0.912 0.908 0.918 0.776 0.836 0.203 F1 0.905 0.925 0.889 0.797 0.864 0.235 Overall accuracy 0.881 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_6_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9273 0.0239 0.0187 0.0149 0.0003 0.0149 building 0.0486 0.9175 0.0183 0.0020 0.0011 0.0126 low_veg 0.0183 0.0038 0.9341 0.0427 0.0000 0.0010 tree 0.0748 0.0036 0.1134 0.8050 0.0018 0.0014 car 0.0139 0.0347 0.0003 0.0047 0.8902 0.0562 clutter 0.4180 0.0565 0.2886 0.0325 0.0056 0.1989 Precision/Correctness 0.914 0.951 0.880 0.847 0.929 0.272 Recall/Completeness 0.927 0.917 0.934 0.805 0.890 0.199 F1 0.921 0.934 0.906 0.826 0.909 0.230 Overall accuracy 0.899 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_6_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9301 0.0205 0.0260 0.0120 0.0044 0.0069 building 0.0554 0.9146 0.0093 0.0045 0.0004 0.0157 low_veg 0.0502 0.0137 0.8470 0.0826 0.0005 0.0062 tree 0.0622 0.0079 0.1307 0.7890 0.0069 0.0034 car 0.0590 0.0114 0.0029 0.0087 0.9031 0.0149 clutter 0.2425 0.2693 0.2494 0.0302 0.0152 0.1935 Precision/Correctness 0.906 0.940 0.807 0.791 0.906 0.328 Recall/Completeness 0.930 0.915 0.847 0.789 0.903 0.193 F1 0.918 0.927 0.826 0.790 0.905 0.243 Overall accuracy 0.881 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_6_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9451 0.0179 0.0212 0.0094 0.0003 0.0061 building 0.0509 0.9234 0.0066 0.0041 0.0003 0.0148 low_veg 0.0349 0.0100 0.8738 0.0770 0.0000 0.0043 tree 0.0508 0.0067 0.1187 0.8150 0.0056 0.0033 car 0.0081 0.0119 0.0012 0.0061 0.9576 0.0151 clutter 0.2292 0.2956 0.2479 0.0271 0.0136 0.1865 Precision/Correctness 0.923 0.949 0.829 0.818 0.958 0.316 Recall/Completeness 0.945 0.923 0.874 0.815 0.958 0.186 F1 0.934 0.936 0.851 0.816 0.958 0.235 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_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8406 0.0768 0.0405 0.0222 0.0052 0.0148 building 0.0171 0.9250 0.0441 0.0060 0.0003 0.0075 low_veg 0.0625 0.0121 0.8034 0.1075 0.0012 0.0133 tree 0.0404 0.0084 0.1603 0.7812 0.0065 0.0032 car 0.0597 0.0136 0.0024 0.0160 0.8924 0.0161 clutter 0.2077 0.1020 0.2394 0.0252 0.0129 0.4129 Precision/Correctness 0.904 0.847 0.648 0.814 0.872 0.706 Recall/Completeness 0.841 0.925 0.803 0.781 0.892 0.413 F1 0.871 0.884 0.717 0.797 0.882 0.521 Overall accuracy 0.821 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 0.8574 0.0763 0.0340 0.0178 0.0006 0.0138 building 0.0141 0.9319 0.0424 0.0049 0.0003 0.0065 low_veg 0.0478 0.0096 0.8306 0.1001 0.0002 0.0117 tree 0.0323 0.0065 0.1463 0.8067 0.0056 0.0027 car 0.0109 0.0139 0.0003 0.0122 0.9462 0.0165 clutter 0.2044 0.1023 0.2362 0.0220 0.0121 0.4231 Precision/Correctness 0.922 0.857 0.660 0.842 0.925 0.730 Recall/Completeness 0.857 0.932 0.831 0.807 0.946 0.423 F1 0.888 0.893 0.735 0.824 0.936 0.536 Overall accuracy 0.840 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 0.8830 0.0494 0.0132 0.0333 0.0022 0.0189 building 0.0644 0.8921 0.0047 0.0064 0.0041 0.0282 low_veg 0.0640 0.0060 0.7356 0.1880 0.0003 0.0060 tree 0.0352 0.0032 0.1362 0.8234 0.0010 0.0009 car 0.0520 0.0106 0.0017 0.0167 0.9123 0.0068 clutter 0.2248 0.0149 0.1512 0.0032 0.0126 0.5933 Precision/Correctness 0.908 0.881 0.700 0.631 0.793 0.765 Recall/Completeness 0.883 0.892 0.736 0.823 0.912 0.593 F1 0.895 0.887 0.717 0.714 0.848 0.668 Overall accuracy 0.840 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 0.8960 0.0472 0.0110 0.0268 0.0007 0.0183 building 0.0621 0.8973 0.0038 0.0053 0.0038 0.0276 low_veg 0.0516 0.0027 0.7585 0.1820 0.0000 0.0052 tree 0.0250 0.0019 0.1312 0.8408 0.0006 0.0006 car 0.0068 0.0118 0.0006 0.0112 0.9621 0.0075 clutter 0.2173 0.0132 0.1519 0.0028 0.0114 0.6034 Precision/Correctness 0.916 0.890 0.709 0.655 0.817 0.775 Recall/Completeness 0.896 0.897 0.759 0.841 0.962 0.603 F1 0.906 0.894 0.733 0.736 0.883 0.679 Overall accuracy 0.853 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image