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 0.913 0.020 0.045 0.014 0.003 0.005 building 0.023 0.958 0.012 0.004 0.000 0.003 low_veg 0.062 0.013 0.853 0.066 0.000 0.006 tree 0.060 0.008 0.168 0.758 0.004 0.003 car 0.107 0.016 0.005 0.010 0.855 0.007 clutter 0.379 0.125 0.183 0.029 0.009 0.275 Precision/Correctness 0.856 0.933 0.769 0.862 0.886 0.746 Recall/Completeness 0.913 0.958 0.853 0.758 0.855 0.275 F1 0.884 0.946 0.809 0.807 0.870 0.401 Overall accuracy 0.854	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 0.933 0.015 0.037 0.010 0.000 0.005 building 0.020 0.965 0.009 0.003 0.000 0.002 low_veg 0.048 0.007 0.882 0.059 0.000 0.005 tree 0.049 0.007 0.152 0.786 0.003 0.003 car 0.035 0.018 0.003 0.006 0.930 0.008 clutter 0.385 0.108 0.172 0.027 0.009 0.299 Precision/Correctness 0.880 0.952 0.794 0.886 0.931 0.776 Recall/Completeness 0.933 0.965 0.882 0.786 0.930 0.299 F1 0.906 0.959 0.835 0.833 0.931 0.431 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 0.9016 0.0214 0.0470 0.0133 0.0023 0.0145 building 0.0170 0.9473 0.0196 0.0126 0.0002 0.0033 low_veg 0.0475 0.0110 0.8666 0.0579 0.0002 0.0168 tree 0.0516 0.0142 0.1628 0.7630 0.0020 0.0064 car 0.1466 0.0151 0.0048 0.0029 0.8103 0.0203 clutter 0.2028 0.2067 0.1556 0.0341 0.0171 0.3836 Precision/Correctness 0.863 0.916 0.824 0.882 0.867 0.402 Recall/Completeness 0.902 0.947 0.867 0.763 0.810 0.384 F1 0.882 0.931 0.845 0.818 0.838 0.393 Overall accuracy 0.856 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 0.9267 0.0142 0.0368 0.0094 0.0003 0.0125 building 0.0129 0.9555 0.0171 0.0119 0.0002 0.0025 low_veg 0.0379 0.0055 0.8926 0.0481 0.0001 0.0159 tree 0.0427 0.0126 0.1411 0.7963 0.0015 0.0058 car 0.0747 0.0166 0.0018 0.0013 0.8817 0.0239 clutter 0.1708 0.2103 0.1215 0.0268 0.0198 0.4508 Precision/Correctness 0.894 0.942 0.852 0.905 0.916 0.393 Recall/Completeness 0.927 0.955 0.893 0.796 0.882 0.451 F1 0.910 0.949 0.872 0.847 0.899 0.420 Overall accuracy 0.884 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 0.8928 0.0105 0.0785 0.0143 0.0020 0.0019 building 0.0104 0.9671 0.0096 0.0105 0.0001 0.0022 low_veg 0.0333 0.0050 0.9084 0.0502 0.0001 0.0031 tree 0.0251 0.0033 0.1755 0.7945 0.0010 0.0006 car 0.0899 0.0364 0.0042 0.0077 0.8552 0.0067 clutter 0.1835 0.2648 0.1805 0.0821 0.0159 0.2732 Precision/Correctness 0.848 0.917 0.832 0.922 0.870 0.332 Recall/Completeness 0.893 0.967 0.908 0.794 0.855 0.273 F1 0.870 0.942 0.869 0.853 0.862 0.300 Overall accuracy 0.867 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 0.9226 0.0061 0.0608 0.0091 0.0002 0.0012 building 0.0071 0.9781 0.0054 0.0076 0.0001 0.0017 low_veg 0.0266 0.0026 0.9261 0.0418 0.0000 0.0029 tree 0.0211 0.0027 0.1594 0.8155 0.0008 0.0005 car 0.0325 0.0431 0.0004 0.0043 0.9129 0.0069 clutter 0.1661 0.2947 0.1314 0.0943 0.0155 0.2980 Precision/Correctness 0.874 0.946 0.851 0.938 0.922 0.305 Recall/Completeness 0.923 0.978 0.926 0.815 0.913 0.298 F1 0.898 0.962 0.887 0.872 0.918 0.302 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 0.8823 0.0223 0.0697 0.0167 0.0037 0.0053 building 0.0178 0.9700 0.0059 0.0039 0.0001 0.0023 low_veg 0.0644 0.0236 0.8371 0.0679 0.0002 0.0068 tree 0.0502 0.0111 0.1339 0.7984 0.0035 0.0029 car 0.1126 0.0197 0.0077 0.0131 0.8366 0.0104 clutter 0.3537 0.1417 0.2740 0.0332 0.0125 0.1850 Precision/Correctness 0.843 0.932 0.774 0.894 0.873 0.503 Recall/Completeness 0.882 0.970 0.837 0.798 0.837 0.185 F1 0.862 0.951 0.804 0.844 0.854 0.271 Overall accuracy 0.857 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 0.9108 0.0147 0.0582 0.0114 0.0006 0.0042 building 0.0141 0.9781 0.0032 0.0029 0.0000 0.0016 low_veg 0.0523 0.0126 0.8720 0.0578 0.0001 0.0052 tree 0.0420 0.0094 0.1167 0.8263 0.0030 0.0026 car 0.0342 0.0235 0.0057 0.0088 0.9179 0.0099 clutter 0.3720 0.1338 0.2456 0.0283 0.0145 0.2057 Precision/Correctness 0.874 0.957 0.805 0.919 0.917 0.528 Recall/Completeness 0.911 0.978 0.872 0.826 0.918 0.206 F1 0.892 0.967 0.837 0.870 0.918 0.296 Overall accuracy 0.887 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 0.8766 0.0297 0.0677 0.0188 0.0033 0.0038 building 0.0188 0.9705 0.0060 0.0030 0.0001 0.0017 low_veg 0.0435 0.0148 0.8849 0.0538 0.0001 0.0029 tree 0.0377 0.0082 0.2426 0.7067 0.0018 0.0030 car 0.1302 0.0408 0.0056 0.0205 0.7948 0.0081 clutter 0.2703 0.1699 0.1944 0.0403 0.0091 0.3161 Precision/Correctness 0.839 0.922 0.740 0.898 0.866 0.771 Recall/Completeness 0.877 0.971 0.885 0.707 0.795 0.316 F1 0.857 0.946 0.806 0.791 0.829 0.448 Overall accuracy 0.838 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 0.9078 0.0217 0.0538 0.0131 0.0006 0.0030 building 0.0158 0.9775 0.0033 0.0021 0.0000 0.0012 low_veg 0.0315 0.0081 0.9138 0.0446 0.0000 0.0020 tree 0.0322 0.0070 0.2271 0.7295 0.0016 0.0026 car 0.0613 0.0446 0.0030 0.0128 0.8710 0.0073 clutter 0.2664 0.1581 0.1583 0.0355 0.0102 0.3717 Precision/Correctness 0.870 0.946 0.763 0.922 0.912 0.814 Recall/Completeness 0.908 0.978 0.914 0.730 0.871 0.372 F1 0.889 0.961 0.832 0.815 0.891 0.510 Overall accuracy 0.865 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 0.9278 0.0160 0.0338 0.0186 0.0020 0.0017 building 0.0169 0.9714 0.0067 0.0034 0.0001 0.0014 low_veg 0.0729 0.0157 0.8063 0.1027 0.0003 0.0021 tree 0.0684 0.0071 0.0913 0.8299 0.0025 0.0008 car 0.1353 0.0096 0.0068 0.0213 0.8228 0.0041 clutter 0.5649 0.1278 0.1734 0.0718 0.0053 0.0568 Precision/Correctness 0.809 0.950 0.764 0.797 0.928 0.723 Recall/Completeness 0.928 0.971 0.806 0.830 0.823 0.057 F1 0.865 0.961 0.785 0.813 0.872 0.105 Overall accuracy 0.851 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 0.9480 0.0099 0.0270 0.0138 0.0002 0.0011 building 0.0119 0.9804 0.0040 0.0025 0.0001 0.0011 low_veg 0.0554 0.0068 0.8419 0.0945 0.0000 0.0015 tree 0.0540 0.0049 0.0758 0.8628 0.0018 0.0006 car 0.0463 0.0093 0.0022 0.0143 0.9236 0.0043 clutter 0.5940 0.1099 0.1655 0.0670 0.0056 0.0581 Precision/Correctness 0.838 0.968 0.799 0.830 0.960 0.764 Recall/Completeness 0.948 0.980 0.842 0.863 0.924 0.058 F1 0.890 0.974 0.820 0.846 0.941 0.108 Overall accuracy 0.879 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 0.8864 0.0211 0.0666 0.0093 0.0048 0.0119 building 0.0161 0.9713 0.0051 0.0043 0.0001 0.0031 low_veg 0.1150 0.0168 0.7634 0.0926 0.0005 0.0117 tree 0.0664 0.0085 0.1642 0.7537 0.0039 0.0033 car 0.0976 0.0132 0.0045 0.0137 0.8581 0.0130 clutter 0.5662 0.0492 0.1481 0.0166 0.0027 0.2171 Precision/Correctness 0.692 0.925 0.734 0.843 0.893 0.748 Recall/Completeness 0.886 0.971 0.763 0.754 0.858 0.217 F1 0.777 0.947 0.748 0.796 0.875 0.337 Overall accuracy 0.783 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 0.9054 0.0157 0.0606 0.0065 0.0007 0.0110 building 0.0125 0.9783 0.0033 0.0034 0.0001 0.0025 low_veg 0.0978 0.0103 0.7932 0.0879 0.0002 0.0105 tree 0.0559 0.0070 0.1464 0.7848 0.0033 0.0027 car 0.0366 0.0149 0.0024 0.0099 0.9224 0.0137 clutter 0.5828 0.0450 0.1344 0.0132 0.0022 0.2224 Precision/Correctness 0.707 0.943 0.756 0.863 0.941 0.776 Recall/Completeness 0.905 0.978 0.793 0.785 0.922 0.222 F1 0.794 0.961 0.774 0.822 0.932 0.346 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_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9185 0.0209 0.0511 0.0040 0.0025 0.0031 building 0.0105 0.9738 0.0113 0.0032 0.0001 0.0011 low_veg 0.0837 0.0247 0.8484 0.0391 0.0007 0.0034 tree 0.1088 0.0121 0.2008 0.6628 0.0095 0.0060 car 0.1282 0.0208 0.0059 0.0026 0.8330 0.0096 clutter 0.3596 0.2580 0.1810 0.0308 0.0115 0.1591 Precision/Correctness 0.876 0.936 0.728 0.898 0.886 0.609 Recall/Completeness 0.918 0.974 0.848 0.663 0.833 0.159 F1 0.897 0.955 0.783 0.762 0.859 0.252 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_15_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9373 0.0133 0.0443 0.0026 0.0004 0.0021 building 0.0079 0.9792 0.0095 0.0025 0.0001 0.0008 low_veg 0.0601 0.0142 0.8891 0.0345 0.0001 0.0020 tree 0.0924 0.0104 0.1793 0.7042 0.0081 0.0057 car 0.0417 0.0238 0.0019 0.0007 0.9210 0.0109 clutter 0.3635 0.2644 0.1693 0.0290 0.0098 0.1640 Precision/Correctness 0.906 0.955 0.755 0.920 0.925 0.639 Recall/Completeness 0.937 0.979 0.889 0.704 0.921 0.164 F1 0.922 0.967 0.816 0.798 0.923 0.261 Overall accuracy 0.895 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 0.9091 0.0217 0.0449 0.0188 0.0028 0.0028 building 0.0181 0.9662 0.0094 0.0036 0.0001 0.0025 low_veg 0.0621 0.0138 0.8402 0.0798 0.0003 0.0038 tree 0.0755 0.0059 0.1558 0.7571 0.0037 0.0019 car 0.1162 0.0090 0.0012 0.0046 0.8644 0.0047 clutter 0.2985 0.3102 0.1594 0.0265 0.0265 0.1790 Precision/Correctness 0.886 0.930 0.792 0.799 0.906 0.679 Recall/Completeness 0.909 0.966 0.840 0.757 0.864 0.179 F1 0.897 0.948 0.816 0.778 0.885 0.283 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 0.9293 0.0147 0.0394 0.0146 0.0004 0.0016 building 0.0148 0.9718 0.0081 0.0029 0.0000 0.0023 low_veg 0.0556 0.0081 0.8616 0.0721 0.0001 0.0024 tree 0.0593 0.0043 0.1410 0.7906 0.0032 0.0017 car 0.0325 0.0102 0.0002 0.0031 0.9491 0.0048 clutter 0.2802 0.3130 0.1511 0.0266 0.0313 0.1979 Precision/Correctness 0.912 0.950 0.819 0.826 0.938 0.715 Recall/Completeness 0.929 0.972 0.862 0.791 0.949 0.198 F1 0.920 0.961 0.840 0.808 0.944 0.310 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9275 0.0100 0.0362 0.0154 0.0025 0.0084 building 0.0183 0.9744 0.0044 0.0019 0.0001 0.0010 low_veg 0.0954 0.0095 0.8074 0.0813 0.0005 0.0059 tree 0.0972 0.0082 0.1354 0.7468 0.0058 0.0066 car 0.1133 0.0181 0.0085 0.0059 0.8487 0.0055 clutter 0.2812 0.1961 0.1880 0.0447 0.0086 0.2814 Precision/Correctness 0.794 0.921 0.775 0.840 0.871 0.807 Recall/Completeness 0.927 0.974 0.807 0.747 0.849 0.281 F1 0.856 0.947 0.791 0.791 0.860 0.417 Overall accuracy 0.833 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 0.9460 0.0062 0.0282 0.0120 0.0003 0.0073 building 0.0160 0.9788 0.0028 0.0015 0.0001 0.0008 low_veg 0.0777 0.0047 0.8354 0.0771 0.0001 0.0051 tree 0.0848 0.0069 0.1258 0.7716 0.0049 0.0061 car 0.0346 0.0192 0.0065 0.0033 0.9309 0.0054 clutter 0.2775 0.1674 0.1838 0.0453 0.0085 0.3175 Precision/Correctness 0.822 0.943 0.798 0.860 0.905 0.831 Recall/Completeness 0.946 0.979 0.835 0.772 0.931 0.317 F1 0.880 0.961 0.816 0.813 0.918 0.459 Overall accuracy 0.858 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 0.9167 0.0124 0.0487 0.0148 0.0024 0.0050 building 0.0284 0.9500 0.0136 0.0052 0.0002 0.0027 low_veg 0.0717 0.0128 0.8094 0.0956 0.0002 0.0104 tree 0.0616 0.0060 0.1305 0.7968 0.0038 0.0014 car 0.1224 0.0127 0.0040 0.0081 0.8491 0.0037 clutter 0.3356 0.1454 0.1863 0.0140 0.0061 0.3126 Precision/Correctness 0.877 0.929 0.694 0.807 0.922 0.833 Recall/Completeness 0.917 0.950 0.809 0.797 0.849 0.313 F1 0.897 0.939 0.747 0.802 0.884 0.455 Overall accuracy 0.852 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 0.9334 0.0076 0.0427 0.0116 0.0005 0.0041 building 0.0234 0.9586 0.0117 0.0040 0.0001 0.0022 low_veg 0.0488 0.0071 0.8437 0.0902 0.0000 0.0102 tree 0.0491 0.0040 0.1190 0.8235 0.0032 0.0012 car 0.0407 0.0151 0.0009 0.0064 0.9329 0.0039 clutter 0.3405 0.1083 0.1864 0.0120 0.0062 0.3467 Precision/Correctness 0.900 0.954 0.715 0.834 0.951 0.854 Recall/Completeness 0.933 0.959 0.844 0.824 0.933 0.347 F1 0.917 0.956 0.774 0.829 0.942 0.493 Overall accuracy 0.876 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 0.9279 0.0237 0.0307 0.0128 0.0015 0.0034 building 0.0347 0.9438 0.0173 0.0016 0.0002 0.0023 low_veg 0.0243 0.0073 0.9343 0.0335 0.0000 0.0005 tree 0.1123 0.0036 0.1480 0.7332 0.0018 0.0012 car 0.1095 0.0400 0.0017 0.0026 0.8395 0.0067 clutter 0.5680 0.0489 0.2215 0.0314 0.0041 0.1260 Precision/Correctness 0.896 0.948 0.854 0.863 0.917 0.544 Recall/Completeness 0.928 0.944 0.934 0.733 0.839 0.126 F1 0.912 0.946 0.892 0.793 0.876 0.205 Overall accuracy 0.893 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 0.9454 0.0197 0.0236 0.0086 0.0002 0.0024 building 0.0306 0.9523 0.0141 0.0010 0.0002 0.0018 low_veg 0.0143 0.0035 0.9537 0.0284 0.0000 0.0002 tree 0.0931 0.0023 0.1364 0.7658 0.0015 0.0010 car 0.0312 0.0473 0.0006 0.0018 0.9119 0.0072 clutter 0.5963 0.0461 0.2068 0.0308 0.0036 0.1163 Precision/Correctness 0.916 0.960 0.875 0.892 0.962 0.574 Recall/Completeness 0.945 0.952 0.954 0.766 0.912 0.116 F1 0.930 0.956 0.913 0.824 0.936 0.193 Overall accuracy 0.913 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 0.9384 0.0108 0.0368 0.0076 0.0052 0.0013 building 0.0556 0.9292 0.0086 0.0019 0.0006 0.0042 low_veg 0.0661 0.0118 0.8680 0.0515 0.0007 0.0018 tree 0.0729 0.0083 0.1428 0.7672 0.0082 0.0006 car 0.0829 0.0052 0.0067 0.0072 0.8945 0.0035 clutter 0.3818 0.2054 0.2624 0.0332 0.0221 0.0951 Precision/Correctness 0.890 0.960 0.784 0.854 0.887 0.499 Recall/Completeness 0.938 0.929 0.868 0.767 0.895 0.095 F1 0.914 0.944 0.824 0.808 0.891 0.160 Overall accuracy 0.887 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 0.9573 0.0056 0.0302 0.0055 0.0005 0.0008 building 0.0520 0.9368 0.0058 0.0013 0.0004 0.0037 low_veg 0.0476 0.0054 0.9003 0.0456 0.0001 0.0010 tree 0.0604 0.0067 0.1282 0.7971 0.0070 0.0005 car 0.0229 0.0044 0.0044 0.0043 0.9603 0.0037 clutter 0.3787 0.2204 0.2546 0.0319 0.0207 0.0937 Precision/Correctness 0.909 0.972 0.811 0.881 0.944 0.521 Recall/Completeness 0.957 0.937 0.900 0.797 0.960 0.094 F1 0.933 0.954 0.853 0.837 0.952 0.159 Overall accuracy 0.909 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image
============================================================= Tile top_potsdam_6_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9172 0.0094 0.0518 0.0071 0.0068 0.0077 building 0.0191 0.9332 0.0434 0.0024 0.0002 0.0017 low_veg 0.0807 0.0049 0.8593 0.0444 0.0016 0.0091 tree 0.0549 0.0084 0.2474 0.6770 0.0071 0.0051 car 0.0711 0.0046 0.0064 0.0082 0.9026 0.0072 clutter 0.2401 0.0408 0.2434 0.0165 0.0203 0.4389 Precision/Correctness 0.893 0.964 0.606 0.905 0.842 0.822 Recall/Completeness 0.917 0.933 0.859 0.677 0.903 0.439 F1 0.905 0.948 0.711 0.775 0.871 0.572 Overall accuracy 0.846 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 0.9391 0.0054 0.0429 0.0049 0.0011 0.0067 building 0.0141 0.9419 0.0407 0.0019 0.0001 0.0013 low_veg 0.0617 0.0012 0.8899 0.0390 0.0003 0.0078 tree 0.0453 0.0068 0.2285 0.7084 0.0062 0.0049 car 0.0184 0.0048 0.0023 0.0057 0.9609 0.0080 clutter 0.2334 0.0398 0.2353 0.0153 0.0196 0.4566 Precision/Correctness 0.914 0.975 0.622 0.926 0.901 0.848 Recall/Completeness 0.939 0.942 0.890 0.708 0.961 0.457 F1 0.926 0.958 0.732 0.803 0.930 0.593 Overall accuracy 0.869 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 0.9129 0.0405 0.0150 0.0221 0.0021 0.0074 building 0.0326 0.9496 0.0038 0.0036 0.0007 0.0098 low_veg 0.0546 0.0175 0.7821 0.1422 0.0005 0.0030 tree 0.0334 0.0103 0.1241 0.8309 0.0009 0.0005 car 0.0741 0.0327 0.0028 0.0197 0.8668 0.0040 clutter 0.3178 0.0306 0.1392 0.0111 0.0045 0.4967 Precision/Correctness 0.909 0.892 0.721 0.702 0.872 0.878 Recall/Completeness 0.913 0.950 0.782 0.831 0.867 0.497 F1 0.911 0.920 0.750 0.761 0.869 0.634 Overall accuracy 0.863 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 0.9284 0.0365 0.0117 0.0160 0.0004 0.0070 building 0.0304 0.9535 0.0032 0.0030 0.0006 0.0093 low_veg 0.0405 0.0122 0.8094 0.1351 0.0002 0.0026 tree 0.0225 0.0077 0.1139 0.8551 0.0006 0.0002 car 0.0184 0.0379 0.0010 0.0139 0.9249 0.0039 clutter 0.3105 0.0288 0.1391 0.0111 0.0040 0.5065 Precision/Correctness 0.918 0.905 0.737 0.733 0.927 0.886 Recall/Completeness 0.928 0.953 0.809 0.855 0.925 0.506 F1 0.923 0.928 0.772 0.789 0.926 0.645 Overall accuracy 0.879 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image