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 0.894 0.020 0.036 0.018 0.002 0.030 building 0.021 0.954 0.009 0.005 0.002 0.009 low_veg 0.045 0.015 0.845 0.069 0.000 0.026 tree 0.041 0.008 0.156 0.784 0.004 0.008 car 0.079 0.005 0.002 0.009 0.895 0.010 clutter 0.303 0.110 0.107 0.027 0.021 0.432 Precision/Correctness 0.884 0.935 0.798 0.853 0.864 0.518 Recall/Completeness 0.894 0.954 0.845 0.784 0.895 0.432 F1 0.889 0.944 0.821 0.817 0.879 0.471 Overall accuracy 0.858	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 0.910 0.017 0.030 0.013 0.000 0.029 building 0.018 0.960 0.008 0.004 0.001 0.008 low_veg 0.033 0.010 0.870 0.062 0.000 0.024 tree 0.032 0.006 0.141 0.811 0.003 0.007 car 0.013 0.005 0.001 0.006 0.965 0.010 clutter 0.304 0.093 0.093 0.023 0.021 0.465 Precision/Correctness 0.905 0.951 0.820 0.879 0.887 0.528 Recall/Completeness 0.910 0.960 0.870 0.811 0.965 0.465 F1 0.908 0.956 0.844 0.843 0.924 0.495 Overall accuracy 0.879
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8748 0.0187 0.0468 0.0143 0.0019 0.0436 building 0.0187 0.9399 0.0078 0.0204 0.0004 0.0128 low_veg 0.0316 0.0143 0.8758 0.0441 0.0006 0.0337 tree 0.0369 0.0133 0.1944 0.7424 0.0018 0.0112 car 0.0935 0.0064 0.0025 0.0057 0.8489 0.0430 clutter 0.1411 0.1685 0.1354 0.0181 0.0300 0.5070 Precision/Correctness 0.894 0.919 0.814 0.891 0.854 0.270 Recall/Completeness 0.875 0.940 0.876 0.742 0.849 0.507 F1 0.884 0.929 0.844 0.810 0.852 0.352 Overall accuracy 0.849 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 0.8972 0.0136 0.0368 0.0099 0.0003 0.0422 building 0.0148 0.9488 0.0053 0.0196 0.0003 0.0112 low_veg 0.0248 0.0103 0.8956 0.0362 0.0005 0.0326 tree 0.0293 0.0117 0.1715 0.7763 0.0013 0.0099 car 0.0308 0.0068 0.0017 0.0025 0.9095 0.0486 clutter 0.1072 0.1613 0.0996 0.0121 0.0369 0.5829 Precision/Correctness 0.920 0.940 0.840 0.913 0.886 0.249 Recall/Completeness 0.897 0.949 0.896 0.776 0.909 0.583 F1 0.909 0.945 0.867 0.839 0.898 0.348 Overall accuracy 0.874 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 0.8544 0.0076 0.0912 0.0116 0.0010 0.0342 building 0.0089 0.9677 0.0091 0.0072 0.0002 0.0068 low_veg 0.0279 0.0043 0.9128 0.0363 0.0000 0.0186 tree 0.0201 0.0055 0.2277 0.7364 0.0011 0.0091 car 0.0748 0.0213 0.0054 0.0038 0.8882 0.0065 clutter 0.1082 0.2640 0.1376 0.0487 0.0508 0.3907 Precision/Correctness 0.867 0.918 0.797 0.937 0.876 0.077 Recall/Completeness 0.854 0.968 0.913 0.736 0.888 0.391 F1 0.861 0.942 0.851 0.825 0.882 0.129 Overall accuracy 0.843 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 0.8804 0.0047 0.0729 0.0071 0.0000 0.0350 building 0.0061 0.9782 0.0052 0.0053 0.0001 0.0050 low_veg 0.0237 0.0024 0.9251 0.0303 0.0000 0.0185 tree 0.0168 0.0048 0.2115 0.7575 0.0008 0.0087 car 0.0112 0.0275 0.0032 0.0017 0.9502 0.0063 clutter 0.0696 0.3075 0.1032 0.0526 0.0649 0.4023 Precision/Correctness 0.888 0.941 0.814 0.951 0.903 0.059 Recall/Completeness 0.880 0.978 0.925 0.757 0.950 0.402 F1 0.884 0.959 0.866 0.843 0.926 0.103 Overall accuracy 0.861 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 0.8689 0.0232 0.0528 0.0210 0.0032 0.0308 building 0.0157 0.9689 0.0059 0.0036 0.0001 0.0058 low_veg 0.0561 0.0247 0.8284 0.0666 0.0002 0.0240 tree 0.0274 0.0077 0.1304 0.8246 0.0038 0.0062 car 0.0753 0.0041 0.0011 0.0097 0.9015 0.0083 clutter 0.2931 0.1497 0.2025 0.0297 0.0139 0.3111 Precision/Correctness 0.874 0.934 0.794 0.895 0.882 0.312 Recall/Completeness 0.869 0.969 0.828 0.825 0.902 0.311 F1 0.872 0.951 0.811 0.858 0.892 0.312 Overall accuracy 0.861 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 0.8952 0.0166 0.0436 0.0143 0.0005 0.0299 building 0.0130 0.9753 0.0042 0.0026 0.0000 0.0049 low_veg 0.0453 0.0142 0.8609 0.0571 0.0001 0.0224 tree 0.0209 0.0060 0.1130 0.8515 0.0031 0.0055 car 0.0137 0.0040 0.0003 0.0051 0.9684 0.0085 clutter 0.3058 0.1410 0.1787 0.0213 0.0155 0.3378 Precision/Correctness 0.902 0.957 0.822 0.920 0.920 0.291 Recall/Completeness 0.895 0.975 0.861 0.852 0.968 0.338 F1 0.899 0.966 0.841 0.884 0.944 0.312 Overall accuracy 0.889 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 0.8372 0.0355 0.0654 0.0232 0.0032 0.0355 building 0.0186 0.9663 0.0054 0.0037 0.0003 0.0057 low_veg 0.0294 0.0137 0.8865 0.0548 0.0001 0.0156 tree 0.0274 0.0095 0.2326 0.7162 0.0025 0.0118 car 0.0695 0.0129 0.0037 0.0182 0.8817 0.0141 clutter 0.2117 0.1620 0.1656 0.0270 0.0103 0.4235 Precision/Correctness 0.870 0.919 0.750 0.895 0.861 0.442 Recall/Completeness 0.837 0.966 0.886 0.716 0.882 0.423 F1 0.853 0.942 0.813 0.796 0.871 0.433 Overall accuracy 0.836 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 0.8640 0.0300 0.0538 0.0168 0.0008 0.0347 building 0.0160 0.9732 0.0031 0.0028 0.0002 0.0047 low_veg 0.0198 0.0080 0.9127 0.0457 0.0000 0.0137 tree 0.0229 0.0082 0.2178 0.7379 0.0023 0.0110 car 0.0097 0.0146 0.0016 0.0114 0.9484 0.0143 clutter 0.2044 0.1515 0.1354 0.0220 0.0120 0.4746 Precision/Correctness 0.898 0.939 0.770 0.920 0.890 0.448 Recall/Completeness 0.864 0.973 0.913 0.738 0.948 0.475 F1 0.881 0.956 0.836 0.819 0.919 0.461 Overall accuracy 0.860 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_4_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9118 0.0187 0.0311 0.0249 0.0012 0.0123 building 0.0150 0.9661 0.0062 0.0055 0.0015 0.0058 low_veg 0.0449 0.0177 0.8029 0.1220 0.0002 0.0123 tree 0.0424 0.0042 0.0627 0.8820 0.0034 0.0053 car 0.0809 0.0030 0.0018 0.0128 0.8980 0.0037 clutter 0.4194 0.1251 0.1508 0.0644 0.0133 0.2271 Precision/Correctness 0.854 0.949 0.797 0.781 0.897 0.652 Recall/Completeness 0.912 0.966 0.803 0.882 0.898 0.227 F1 0.882 0.957 0.800 0.828 0.897 0.337 Overall accuracy 0.864 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_4_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9283 0.0156 0.0260 0.0185 0.0001 0.0114 building 0.0102 0.9753 0.0039 0.0043 0.0014 0.0049 low_veg 0.0317 0.0118 0.8323 0.1133 0.0000 0.0109 tree 0.0317 0.0025 0.0493 0.9092 0.0025 0.0047 car 0.0099 0.0027 0.0002 0.0070 0.9767 0.0036 clutter 0.4264 0.1112 0.1413 0.0579 0.0141 0.2491 Precision/Correctness 0.881 0.962 0.827 0.814 0.911 0.673 Recall/Completeness 0.928 0.975 0.832 0.909 0.977 0.249 F1 0.904 0.969 0.830 0.859 0.943 0.364 Overall accuracy 0.888 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 0.8712 0.0211 0.0513 0.0144 0.0029 0.0392 building 0.0137 0.9711 0.0024 0.0057 0.0017 0.0053 low_veg 0.0811 0.0185 0.7726 0.0705 0.0006 0.0568 tree 0.0415 0.0071 0.1401 0.7964 0.0044 0.0104 car 0.0742 0.0055 0.0030 0.0124 0.8916 0.0133 clutter 0.4963 0.0414 0.0884 0.0112 0.0030 0.3597 Precision/Correctness 0.735 0.928 0.785 0.870 0.896 0.574 Recall/Completeness 0.871 0.971 0.773 0.796 0.892 0.360 F1 0.797 0.949 0.779 0.832 0.894 0.442 Overall accuracy 0.804 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 0.8868 0.0164 0.0472 0.0103 0.0003 0.0390 building 0.0108 0.9780 0.0009 0.0046 0.0015 0.0042 low_veg 0.0669 0.0128 0.7998 0.0631 0.0003 0.0571 tree 0.0327 0.0054 0.1221 0.8269 0.0037 0.0093 car 0.0177 0.0058 0.0029 0.0088 0.9516 0.0132 clutter 0.5101 0.0372 0.0755 0.0081 0.0022 0.3669 Precision/Correctness 0.748 0.946 0.807 0.894 0.928 0.591 Recall/Completeness 0.887 0.978 0.800 0.827 0.952 0.367 F1 0.812 0.962 0.803 0.859 0.940 0.453 Overall accuracy 0.824 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 0.8819 0.0233 0.0450 0.0089 0.0021 0.0387 building 0.0076 0.9720 0.0114 0.0050 0.0006 0.0035 low_veg 0.0545 0.0247 0.8501 0.0447 0.0005 0.0255 tree 0.0750 0.0075 0.1551 0.7433 0.0109 0.0082 car 0.0894 0.0033 0.0023 0.0021 0.8927 0.0102 clutter 0.3382 0.2464 0.0889 0.0209 0.0129 0.2929 Precision/Correctness 0.902 0.937 0.771 0.884 0.885 0.300 Recall/Completeness 0.882 0.972 0.850 0.743 0.893 0.293 F1 0.892 0.954 0.809 0.807 0.889 0.296 Overall accuracy 0.870 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 0.8962 0.0176 0.0411 0.0063 0.0003 0.0384 building 0.0050 0.9778 0.0099 0.0043 0.0005 0.0025 low_veg 0.0343 0.0161 0.8872 0.0385 0.0001 0.0239 tree 0.0611 0.0056 0.1337 0.7833 0.0092 0.0071 car 0.0181 0.0037 0.0002 0.0007 0.9658 0.0116 clutter 0.3550 0.2512 0.0729 0.0164 0.0115 0.2930 Precision/Correctness 0.929 0.954 0.794 0.909 0.914 0.271 Recall/Completeness 0.896 0.978 0.887 0.783 0.966 0.293 F1 0.912 0.966 0.838 0.841 0.939 0.282 Overall accuracy 0.894 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 0.8922 0.0256 0.0321 0.0228 0.0016 0.0257 building 0.0167 0.9578 0.0050 0.0050 0.0010 0.0145 low_veg 0.0451 0.0210 0.8268 0.0802 0.0001 0.0267 tree 0.0527 0.0059 0.1331 0.8002 0.0037 0.0044 car 0.0965 0.0055 0.0004 0.0056 0.8875 0.0044 clutter 0.2388 0.2339 0.1136 0.0289 0.0720 0.3128 Precision/Correctness 0.908 0.929 0.831 0.795 0.863 0.353 Recall/Completeness 0.892 0.958 0.827 0.800 0.887 0.313 F1 0.900 0.943 0.829 0.798 0.875 0.332 Overall accuracy 0.868 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 0.9101 0.0205 0.0281 0.0173 0.0001 0.0238 building 0.0136 0.9635 0.0040 0.0041 0.0009 0.0140 low_veg 0.0383 0.0160 0.8501 0.0718 0.0000 0.0238 tree 0.0385 0.0040 0.1195 0.8312 0.0032 0.0037 car 0.0155 0.0064 0.0000 0.0033 0.9703 0.0046 clutter 0.2212 0.2203 0.1017 0.0284 0.0861 0.3423 Precision/Correctness 0.933 0.947 0.855 0.824 0.879 0.334 Recall/Completeness 0.910 0.963 0.850 0.831 0.970 0.342 F1 0.921 0.955 0.852 0.828 0.922 0.338 Overall accuracy 0.892 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 0.9053 0.0118 0.0278 0.0272 0.0013 0.0266 building 0.0191 0.9682 0.0034 0.0021 0.0006 0.0065 low_veg 0.0661 0.0107 0.7915 0.1024 0.0006 0.0287 tree 0.0550 0.0075 0.1013 0.8176 0.0068 0.0119 car 0.0993 0.0098 0.0047 0.0062 0.8631 0.0169 clutter 0.2171 0.1660 0.0930 0.0530 0.0189 0.4520 Precision/Correctness 0.840 0.927 0.832 0.812 0.840 0.662 Recall/Completeness 0.905 0.968 0.791 0.818 0.863 0.452 F1 0.872 0.947 0.811 0.815 0.851 0.537 Overall accuracy 0.847 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 0.9207 0.0094 0.0225 0.0217 0.0002 0.0255 building 0.0166 0.9733 0.0021 0.0016 0.0006 0.0059 low_veg 0.0496 0.0076 0.8178 0.0974 0.0002 0.0273 tree 0.0449 0.0063 0.0925 0.8399 0.0057 0.0107 car 0.0137 0.0108 0.0049 0.0033 0.9487 0.0186 clutter 0.2023 0.1362 0.0812 0.0517 0.0191 0.5095 Precision/Correctness 0.871 0.946 0.854 0.835 0.857 0.681 Recall/Completeness 0.921 0.973 0.818 0.840 0.949 0.510 F1 0.895 0.960 0.835 0.837 0.900 0.583 Overall accuracy 0.872 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 0.8951 0.0079 0.0390 0.0207 0.0014 0.0359 building 0.0173 0.9589 0.0084 0.0060 0.0018 0.0077 low_veg 0.0527 0.0098 0.7685 0.1333 0.0001 0.0355 tree 0.0419 0.0051 0.0903 0.8527 0.0047 0.0053 car 0.1103 0.0010 0.0019 0.0105 0.8636 0.0127 clutter 0.3671 0.1108 0.0786 0.0251 0.0315 0.3869 Precision/Correctness 0.888 0.948 0.765 0.762 0.858 0.549 Recall/Completeness 0.895 0.959 0.769 0.853 0.864 0.387 F1 0.891 0.954 0.767 0.805 0.861 0.454 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_15_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9072 0.0047 0.0358 0.0163 0.0006 0.0353 building 0.0131 0.9669 0.0075 0.0047 0.0014 0.0064 low_veg 0.0330 0.0058 0.7977 0.1285 0.0000 0.0350 tree 0.0315 0.0033 0.0808 0.8758 0.0040 0.0046 car 0.0193 0.0011 0.0000 0.0086 0.9573 0.0137 clutter 0.3798 0.0768 0.0704 0.0238 0.0343 0.4150 Precision/Correctness 0.908 0.968 0.784 0.789 0.860 0.558 Recall/Completeness 0.907 0.967 0.798 0.876 0.957 0.415 F1 0.908 0.968 0.791 0.830 0.906 0.476 Overall accuracy 0.873 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 0.9294 0.0245 0.0202 0.0125 0.0022 0.0111 building 0.0324 0.9408 0.0169 0.0044 0.0007 0.0047 low_veg 0.0226 0.0106 0.9293 0.0344 0.0000 0.0031 tree 0.0987 0.0054 0.1127 0.7794 0.0015 0.0024 car 0.0719 0.0023 0.0002 0.0014 0.9206 0.0037 clutter 0.3864 0.0764 0.2207 0.0262 0.0415 0.2488 Precision/Correctness 0.912 0.942 0.881 0.865 0.847 0.444 Recall/Completeness 0.929 0.941 0.929 0.779 0.921 0.249 F1 0.921 0.941 0.905 0.820 0.882 0.319 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_13_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9418 0.0216 0.0172 0.0080 0.0013 0.0101 building 0.0288 0.9480 0.0147 0.0039 0.0006 0.0040 low_veg 0.0124 0.0079 0.9486 0.0287 0.0000 0.0024 tree 0.0791 0.0042 0.1009 0.8128 0.0011 0.0020 car 0.0123 0.0028 0.0001 0.0004 0.9818 0.0027 clutter 0.3931 0.0729 0.2186 0.0234 0.0448 0.2473 Precision/Correctness 0.931 0.952 0.897 0.894 0.862 0.445 Recall/Completeness 0.942 0.948 0.949 0.813 0.982 0.247 F1 0.936 0.950 0.922 0.851 0.918 0.318 Overall accuracy 0.918 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 0.9378 0.0074 0.0264 0.0116 0.0034 0.0134 building 0.0533 0.9241 0.0033 0.0016 0.0021 0.0156 low_veg 0.0538 0.0196 0.8369 0.0766 0.0006 0.0125 tree 0.0547 0.0103 0.1087 0.8157 0.0077 0.0028 car 0.0540 0.0028 0.0011 0.0094 0.9285 0.0041 clutter 0.3067 0.2220 0.1925 0.0323 0.0251 0.2214 Precision/Correctness 0.905 0.958 0.831 0.812 0.897 0.295 Recall/Completeness 0.938 0.924 0.837 0.816 0.929 0.221 F1 0.921 0.941 0.834 0.814 0.912 0.253 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_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9509 0.0046 0.0226 0.0088 0.0001 0.0130 building 0.0490 0.9316 0.0017 0.0011 0.0017 0.0150 low_veg 0.0356 0.0154 0.8676 0.0702 0.0001 0.0111 tree 0.0442 0.0085 0.0964 0.8420 0.0064 0.0025 car 0.0039 0.0020 0.0002 0.0061 0.9835 0.0043 clutter 0.2996 0.2416 0.1906 0.0281 0.0241 0.2160 Precision/Correctness 0.923 0.967 0.852 0.841 0.935 0.276 Recall/Completeness 0.951 0.932 0.868 0.842 0.983 0.216 F1 0.937 0.949 0.860 0.842 0.959 0.242 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_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9091 0.0097 0.0269 0.0129 0.0038 0.0376 building 0.0177 0.9220 0.0426 0.0057 0.0015 0.0106 low_veg 0.0497 0.0153 0.7979 0.0770 0.0012 0.0589 tree 0.0399 0.0057 0.1434 0.8019 0.0053 0.0037 car 0.0638 0.0018 0.0018 0.0114 0.9130 0.0083 clutter 0.1319 0.0190 0.1010 0.0185 0.0315 0.6981 Precision/Correctness 0.926 0.964 0.710 0.867 0.856 0.614 Recall/Completeness 0.909 0.922 0.798 0.802 0.913 0.698 F1 0.917 0.942 0.751 0.833 0.884 0.653 Overall accuracy 0.867 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 0.9239 0.0077 0.0222 0.0093 0.0002 0.0367 building 0.0134 0.9299 0.0417 0.0047 0.0013 0.0090 low_veg 0.0296 0.0137 0.8296 0.0678 0.0002 0.0591 tree 0.0301 0.0041 0.1287 0.8299 0.0043 0.0030 car 0.0086 0.0020 0.0001 0.0080 0.9722 0.0091 clutter 0.1244 0.0176 0.0945 0.0150 0.0304 0.7181 Precision/Correctness 0.947 0.972 0.725 0.894 0.893 0.630 Recall/Completeness 0.924 0.930 0.830 0.830 0.972 0.718 F1 0.935 0.950 0.774 0.861 0.931 0.671 Overall accuracy 0.887 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 0.8796 0.0425 0.0109 0.0268 0.0018 0.0386 building 0.0331 0.9286 0.0026 0.0019 0.0089 0.0248 low_veg 0.0552 0.0166 0.7542 0.1642 0.0004 0.0094 tree 0.0281 0.0088 0.1260 0.8329 0.0016 0.0026 car 0.0536 0.0025 0.0015 0.0160 0.9257 0.0007 clutter 0.2026 0.0286 0.0360 0.0023 0.0146 0.7159 Precision/Correctness 0.927 0.889 0.800 0.678 0.744 0.714 Recall/Completeness 0.880 0.929 0.754 0.833 0.926 0.716 F1 0.903 0.908 0.776 0.748 0.825 0.715 Overall accuracy 0.860 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 0.8922 0.0390 0.0088 0.0203 0.0007 0.0390 building 0.0312 0.9321 0.0020 0.0013 0.0088 0.0246 low_veg 0.0419 0.0124 0.7808 0.1562 0.0001 0.0085 tree 0.0182 0.0068 0.1176 0.8544 0.0011 0.0019 car 0.0047 0.0025 0.0006 0.0110 0.9806 0.0006 clutter 0.1948 0.0262 0.0343 0.0019 0.0134 0.7294 Precision/Correctness 0.936 0.900 0.819 0.709 0.741 0.720 Recall/Completeness 0.892 0.932 0.781 0.854 0.981 0.729 F1 0.914 0.916 0.799 0.775 0.844 0.725 Overall accuracy 0.874 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image