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 0.905 0.014 0.043 0.015 0.003 0.020 building 0.019 0.953 0.012 0.003 0.001 0.013 low_veg 0.044 0.009 0.879 0.054 0.000 0.015 tree 0.042 0.006 0.149 0.796 0.002 0.004 car 0.070 0.005 0.004 0.015 0.888 0.018 clutter 0.250 0.103 0.175 0.020 0.011 0.441 Precision/Correctness 0.894 0.950 0.788 0.884 0.898 0.599 Recall/Completeness 0.905 0.953 0.879 0.796 0.888 0.441 F1 0.899 0.951 0.831 0.838 0.893 0.508 Overall accuracy 0.871	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 0.925 0.010 0.036 0.010 0.000 0.019 building 0.016 0.961 0.009 0.002 0.000 0.011 low_veg 0.032 0.005 0.904 0.046 0.000 0.013 tree 0.033 0.005 0.133 0.824 0.002 0.004 car 0.013 0.006 0.002 0.009 0.952 0.018 clutter 0.246 0.088 0.163 0.016 0.011 0.476 Precision/Correctness 0.916 0.965 0.812 0.910 0.938 0.619 Recall/Completeness 0.925 0.961 0.904 0.824 0.952 0.476 F1 0.920 0.963 0.855 0.865 0.945 0.538 Overall accuracy 0.894
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8786 0.0168 0.0520 0.0120 0.0024 0.0383 building 0.0200 0.9322 0.0220 0.0054 0.0001 0.0203 low_veg 0.0258 0.0071 0.8816 0.0425 0.0001 0.0429 tree 0.0403 0.0102 0.1436 0.7960 0.0013 0.0085 car 0.0861 0.0062 0.0041 0.0087 0.8540 0.0408 clutter 0.1337 0.1583 0.1397 0.0273 0.0248 0.5161 Precision/Correctness 0.898 0.936 0.835 0.914 0.875 0.258 Recall/Completeness 0.879 0.932 0.882 0.796 0.854 0.516 F1 0.888 0.934 0.857 0.851 0.864 0.344 Overall accuracy 0.863 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 0.9021 0.0116 0.0430 0.0072 0.0002 0.0358 building 0.0167 0.9429 0.0178 0.0043 0.0001 0.0182 low_veg 0.0177 0.0033 0.9047 0.0312 0.0000 0.0430 tree 0.0319 0.0087 0.1244 0.8266 0.0009 0.0075 car 0.0321 0.0065 0.0018 0.0034 0.9104 0.0458 clutter 0.1016 0.1587 0.1024 0.0192 0.0300 0.5882 Precision/Correctness 0.926 0.957 0.860 0.940 0.923 0.234 Recall/Completeness 0.902 0.943 0.905 0.827 0.910 0.588 F1 0.914 0.950 0.882 0.880 0.917 0.335 Overall accuracy 0.889 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 0.8789 0.0063 0.0914 0.0132 0.0018 0.0083 building 0.0086 0.9692 0.0081 0.0051 0.0000 0.0090 low_veg 0.0130 0.0042 0.9436 0.0347 0.0000 0.0045 tree 0.0169 0.0030 0.1922 0.7856 0.0007 0.0016 car 0.0639 0.0056 0.0085 0.0125 0.8981 0.0116 clutter 0.1034 0.2423 0.1914 0.0536 0.0081 0.4013 Precision/Correctness 0.914 0.935 0.823 0.943 0.904 0.247 Recall/Completeness 0.879 0.969 0.944 0.786 0.898 0.401 F1 0.896 0.952 0.879 0.857 0.901 0.306 Overall accuracy 0.877 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 0.9080 0.0032 0.0732 0.0082 0.0002 0.0071 building 0.0063 0.9804 0.0031 0.0029 0.0000 0.0073 low_veg 0.0082 0.0024 0.9577 0.0276 0.0000 0.0041 tree 0.0132 0.0022 0.1755 0.8071 0.0005 0.0014 car 0.0106 0.0051 0.0084 0.0068 0.9557 0.0133 clutter 0.0730 0.2752 0.1468 0.0561 0.0089 0.4400 Precision/Correctness 0.939 0.958 0.842 0.957 0.951 0.227 Recall/Completeness 0.908 0.980 0.958 0.807 0.956 0.440 F1 0.923 0.969 0.896 0.876 0.953 0.300 Overall accuracy 0.896 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 0.8754 0.0166 0.0550 0.0183 0.0037 0.0309 building 0.0138 0.9636 0.0069 0.0038 0.0000 0.0118 low_veg 0.0548 0.0179 0.8657 0.0488 0.0001 0.0126 tree 0.0341 0.0067 0.1224 0.8311 0.0021 0.0035 car 0.0704 0.0044 0.0021 0.0196 0.8904 0.0131 clutter 0.1930 0.1298 0.2631 0.0285 0.0086 0.3771 Precision/Correctness 0.881 0.948 0.800 0.915 0.903 0.380 Recall/Completeness 0.875 0.964 0.866 0.831 0.890 0.377 F1 0.878 0.956 0.831 0.871 0.896 0.378 Overall accuracy 0.873 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 0.9030 0.0104 0.0451 0.0117 0.0003 0.0295 building 0.0111 0.9724 0.0038 0.0027 0.0000 0.0100 low_veg 0.0433 0.0098 0.8982 0.0390 0.0000 0.0096 tree 0.0261 0.0054 0.1043 0.8597 0.0016 0.0029 car 0.0143 0.0049 0.0012 0.0112 0.9565 0.0119 clutter 0.1798 0.1203 0.2368 0.0222 0.0100 0.4309 Precision/Correctness 0.910 0.968 0.830 0.940 0.954 0.380 Recall/Completeness 0.903 0.972 0.898 0.860 0.956 0.431 F1 0.906 0.970 0.863 0.898 0.955 0.404 Overall accuracy 0.902 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 0.8633 0.0171 0.0771 0.0192 0.0039 0.0194 building 0.0144 0.9718 0.0063 0.0028 0.0001 0.0046 low_veg 0.0299 0.0083 0.9044 0.0478 0.0001 0.0095 tree 0.0264 0.0048 0.2043 0.7582 0.0016 0.0048 car 0.0652 0.0029 0.0024 0.0250 0.8626 0.0418 clutter 0.1844 0.1479 0.1964 0.0388 0.0047 0.4279 Precision/Correctness 0.881 0.949 0.763 0.911 0.881 0.579 Recall/Completeness 0.863 0.972 0.904 0.758 0.863 0.428 F1 0.872 0.960 0.828 0.828 0.871 0.492 Overall accuracy 0.858 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 0.8930 0.0110 0.0653 0.0127 0.0007 0.0172 building 0.0125 0.9791 0.0031 0.0018 0.0000 0.0034 low_veg 0.0196 0.0034 0.9318 0.0376 0.0000 0.0075 tree 0.0213 0.0039 0.1882 0.7812 0.0013 0.0041 car 0.0125 0.0028 0.0006 0.0156 0.9219 0.0466 clutter 0.1739 0.1374 0.1675 0.0332 0.0052 0.4828 Precision/Correctness 0.910 0.967 0.785 0.936 0.932 0.609 Recall/Completeness 0.893 0.979 0.932 0.781 0.922 0.483 F1 0.901 0.973 0.852 0.852 0.927 0.539 Overall accuracy 0.884 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 0.9277 0.0087 0.0343 0.0200 0.0026 0.0067 building 0.0150 0.9680 0.0072 0.0035 0.0001 0.0062 low_veg 0.0571 0.0072 0.8320 0.0960 0.0002 0.0075 tree 0.0394 0.0039 0.0633 0.8880 0.0023 0.0029 car 0.0655 0.0019 0.0025 0.0215 0.9030 0.0056 clutter 0.4612 0.1027 0.1848 0.0522 0.0144 0.1847 Precision/Correctness 0.846 0.966 0.783 0.819 0.902 0.688 Recall/Completeness 0.928 0.968 0.832 0.888 0.903 0.185 F1 0.885 0.967 0.807 0.852 0.903 0.291 Overall accuracy 0.871 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 0.9469 0.0053 0.0281 0.0138 0.0006 0.0053 building 0.0100 0.9791 0.0035 0.0024 0.0001 0.0051 low_veg 0.0426 0.0025 0.8636 0.0856 0.0000 0.0056 tree 0.0263 0.0024 0.0503 0.9171 0.0017 0.0023 car 0.0088 0.0015 0.0008 0.0124 0.9717 0.0047 clutter 0.4772 0.0875 0.1769 0.0445 0.0156 0.1982 Precision/Correctness 0.873 0.978 0.816 0.854 0.927 0.727 Recall/Completeness 0.947 0.979 0.864 0.917 0.972 0.198 F1 0.908 0.979 0.839 0.885 0.949 0.312 Overall accuracy 0.898 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 0.8668 0.0189 0.0627 0.0130 0.0043 0.0343 building 0.0171 0.9652 0.0055 0.0038 0.0001 0.0081 low_veg 0.0743 0.0139 0.8249 0.0543 0.0005 0.0321 tree 0.0392 0.0057 0.1654 0.7792 0.0025 0.0080 car 0.0535 0.0061 0.0075 0.0228 0.8889 0.0212 clutter 0.3633 0.0376 0.1391 0.0108 0.0021 0.4472 Precision/Correctness 0.772 0.938 0.753 0.890 0.913 0.692 Recall/Completeness 0.867 0.965 0.825 0.779 0.889 0.447 F1 0.817 0.951 0.787 0.831 0.901 0.543 Overall accuracy 0.819 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 0.8851 0.0143 0.0581 0.0082 0.0005 0.0337 building 0.0141 0.9725 0.0036 0.0028 0.0001 0.0068 low_veg 0.0601 0.0084 0.8534 0.0469 0.0002 0.0310 tree 0.0304 0.0042 0.1472 0.8094 0.0019 0.0068 car 0.0101 0.0054 0.0080 0.0167 0.9382 0.0216 clutter 0.3719 0.0338 0.1243 0.0081 0.0015 0.4602 Precision/Correctness 0.788 0.955 0.774 0.915 0.960 0.714 Recall/Completeness 0.885 0.973 0.853 0.809 0.938 0.460 F1 0.834 0.963 0.811 0.859 0.949 0.560 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 0.9053 0.0179 0.0575 0.0054 0.0029 0.0110 building 0.0111 0.9703 0.0092 0.0013 0.0004 0.0077 low_veg 0.0675 0.0227 0.8700 0.0276 0.0005 0.0118 tree 0.0850 0.0095 0.1754 0.7164 0.0063 0.0074 car 0.0782 0.0161 0.0030 0.0093 0.8743 0.0192 clutter 0.2208 0.2339 0.2241 0.0106 0.0078 0.3028 Precision/Correctness 0.902 0.943 0.734 0.926 0.904 0.480 Recall/Completeness 0.905 0.970 0.870 0.716 0.874 0.303 F1 0.904 0.956 0.797 0.808 0.889 0.371 Overall accuracy 0.877 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 0.9219 0.0122 0.0530 0.0033 0.0004 0.0093 building 0.0081 0.9773 0.0068 0.0009 0.0004 0.0065 low_veg 0.0450 0.0148 0.9090 0.0216 0.0001 0.0095 tree 0.0684 0.0079 0.1543 0.7578 0.0049 0.0066 car 0.0109 0.0177 0.0004 0.0041 0.9464 0.0206 clutter 0.2071 0.2382 0.2273 0.0075 0.0059 0.3140 Precision/Correctness 0.931 0.959 0.758 0.950 0.946 0.483 Recall/Completeness 0.922 0.977 0.909 0.758 0.946 0.314 F1 0.926 0.968 0.827 0.843 0.946 0.381 Overall accuracy 0.903 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 0.9205 0.0125 0.0348 0.0191 0.0022 0.0108 building 0.0142 0.9704 0.0063 0.0027 0.0000 0.0064 low_veg 0.0694 0.0112 0.8393 0.0687 0.0001 0.0113 tree 0.0575 0.0057 0.1219 0.8091 0.0022 0.0036 car 0.0856 0.0015 0.0007 0.0076 0.8866 0.0180 clutter 0.2193 0.2249 0.1466 0.0177 0.0503 0.3413 Precision/Correctness 0.901 0.951 0.829 0.826 0.895 0.563 Recall/Completeness 0.920 0.970 0.839 0.809 0.887 0.341 F1 0.911 0.961 0.834 0.818 0.891 0.425 Overall accuracy 0.886 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 0.9396 0.0071 0.0305 0.0141 0.0002 0.0084 building 0.0106 0.9770 0.0048 0.0020 0.0000 0.0056 low_veg 0.0626 0.0063 0.8616 0.0603 0.0000 0.0092 tree 0.0414 0.0041 0.1078 0.8418 0.0017 0.0033 car 0.0170 0.0014 0.0000 0.0045 0.9588 0.0183 clutter 0.1888 0.2099 0.1382 0.0157 0.0619 0.3855 Precision/Correctness 0.925 0.969 0.854 0.855 0.919 0.579 Recall/Completeness 0.940 0.977 0.862 0.842 0.959 0.385 F1 0.932 0.973 0.858 0.848 0.938 0.463 Overall accuracy 0.910 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 0.9247 0.0069 0.0294 0.0225 0.0020 0.0144 building 0.0172 0.9691 0.0067 0.0012 0.0002 0.0056 low_veg 0.0586 0.0074 0.8495 0.0720 0.0002 0.0123 tree 0.0546 0.0058 0.1064 0.8238 0.0034 0.0060 car 0.0838 0.0018 0.0117 0.0180 0.8620 0.0227 clutter 0.1738 0.1787 0.1905 0.0331 0.0061 0.4178 Precision/Correctness 0.860 0.932 0.805 0.857 0.910 0.769 Recall/Completeness 0.925 0.969 0.850 0.824 0.862 0.418 F1 0.891 0.950 0.827 0.840 0.885 0.541 Overall accuracy 0.863 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 0.9437 0.0039 0.0227 0.0167 0.0004 0.0127 building 0.0154 0.9742 0.0044 0.0008 0.0001 0.0050 low_veg 0.0429 0.0044 0.8761 0.0665 0.0000 0.0100 tree 0.0436 0.0047 0.0967 0.8473 0.0026 0.0051 car 0.0135 0.0017 0.0106 0.0097 0.9408 0.0237 clutter 0.1583 0.1542 0.1838 0.0307 0.0062 0.4668 Precision/Correctness 0.889 0.951 0.828 0.881 0.938 0.797 Recall/Completeness 0.944 0.974 0.876 0.847 0.941 0.467 F1 0.915 0.962 0.852 0.864 0.939 0.589 Overall accuracy 0.888 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 0.8728 0.0140 0.0486 0.0168 0.0020 0.0458 building 0.0214 0.9381 0.0158 0.0049 0.0004 0.0194 low_veg 0.0497 0.0052 0.8310 0.0974 0.0001 0.0165 tree 0.0420 0.0042 0.1078 0.8410 0.0023 0.0028 car 0.1017 0.0054 0.0023 0.0138 0.8609 0.0159 clutter 0.2518 0.1480 0.1243 0.0149 0.0081 0.4530 Precision/Correctness 0.903 0.930 0.727 0.808 0.929 0.548 Recall/Completeness 0.873 0.938 0.831 0.841 0.861 0.453 F1 0.887 0.934 0.776 0.824 0.894 0.496 Overall accuracy 0.850 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 0.8866 0.0104 0.0439 0.0125 0.0006 0.0461 building 0.0175 0.9474 0.0139 0.0038 0.0003 0.0172 low_veg 0.0312 0.0020 0.8591 0.0914 0.0000 0.0162 tree 0.0313 0.0026 0.0953 0.8667 0.0017 0.0024 car 0.0258 0.0047 0.0002 0.0110 0.9423 0.0160 clutter 0.2498 0.1211 0.1119 0.0123 0.0082 0.4968 Precision/Correctness 0.924 0.950 0.749 0.837 0.951 0.557 Recall/Completeness 0.887 0.947 0.859 0.867 0.942 0.497 F1 0.905 0.949 0.800 0.852 0.947 0.525 Overall accuracy 0.872 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 0.9225 0.0224 0.0289 0.0137 0.0019 0.0105 building 0.0378 0.9355 0.0189 0.0011 0.0016 0.0050 low_veg 0.0201 0.0041 0.9437 0.0292 0.0000 0.0030 tree 0.0793 0.0034 0.1067 0.8079 0.0009 0.0019 car 0.0611 0.0042 0.0005 0.0049 0.9106 0.0186 clutter 0.2578 0.0475 0.3736 0.0216 0.0090 0.2906 Precision/Correctness 0.922 0.954 0.862 0.882 0.888 0.485 Recall/Completeness 0.923 0.935 0.944 0.808 0.911 0.291 F1 0.922 0.945 0.901 0.843 0.899 0.363 Overall accuracy 0.904 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 0.9386 0.0199 0.0236 0.0086 0.0003 0.0089 building 0.0340 0.9455 0.0144 0.0006 0.0014 0.0040 low_veg 0.0112 0.0018 0.9604 0.0243 0.0000 0.0022 tree 0.0611 0.0022 0.0945 0.8401 0.0005 0.0015 car 0.0098 0.0053 0.0002 0.0028 0.9652 0.0168 clutter 0.2456 0.0471 0.3903 0.0203 0.0083 0.2883 Precision/Correctness 0.940 0.963 0.881 0.911 0.933 0.501 Recall/Completeness 0.939 0.945 0.960 0.840 0.965 0.288 F1 0.939 0.954 0.919 0.874 0.949 0.366 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_14_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9453 0.0054 0.0278 0.0083 0.0049 0.0082 building 0.0408 0.9295 0.0083 0.0012 0.0003 0.0199 low_veg 0.0521 0.0067 0.8830 0.0502 0.0005 0.0075 tree 0.0618 0.0083 0.1053 0.8182 0.0050 0.0013 car 0.0517 0.0024 0.0038 0.0124 0.9180 0.0117 clutter 0.3166 0.0831 0.2247 0.0208 0.0231 0.3316 Precision/Correctness 0.913 0.979 0.826 0.864 0.903 0.415 Recall/Completeness 0.945 0.930 0.883 0.818 0.918 0.332 F1 0.929 0.954 0.853 0.840 0.910 0.369 Overall accuracy 0.904 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 0.9613 0.0025 0.0227 0.0056 0.0004 0.0074 building 0.0362 0.9390 0.0049 0.0007 0.0002 0.0190 low_veg 0.0358 0.0035 0.9112 0.0439 0.0001 0.0055 tree 0.0494 0.0069 0.0914 0.8472 0.0039 0.0012 car 0.0051 0.0012 0.0023 0.0078 0.9719 0.0116 clutter 0.3071 0.0892 0.2145 0.0161 0.0222 0.3510 Precision/Correctness 0.931 0.986 0.852 0.894 0.958 0.419 Recall/Completeness 0.961 0.939 0.911 0.847 0.972 0.351 F1 0.946 0.962 0.880 0.870 0.965 0.382 Overall accuracy 0.924 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 0.9163 0.0074 0.0406 0.0078 0.0049 0.0229 building 0.0123 0.9311 0.0414 0.0013 0.0032 0.0107 low_veg 0.0537 0.0048 0.8883 0.0375 0.0010 0.0147 tree 0.0523 0.0062 0.1943 0.7376 0.0045 0.0051 car 0.0593 0.0015 0.0039 0.0094 0.9144 0.0114 clutter 0.1378 0.0341 0.2358 0.0065 0.0108 0.5751 Precision/Correctness 0.923 0.970 0.655 0.924 0.872 0.704 Recall/Completeness 0.916 0.931 0.888 0.738 0.914 0.575 F1 0.919 0.950 0.754 0.820 0.893 0.633 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 0.9349 0.0050 0.0330 0.0048 0.0004 0.0218 building 0.0075 0.9405 0.0386 0.0007 0.0031 0.0095 low_veg 0.0367 0.0027 0.9171 0.0302 0.0001 0.0131 tree 0.0428 0.0048 0.1762 0.7683 0.0035 0.0044 car 0.0092 0.0013 0.0005 0.0052 0.9724 0.0115 clutter 0.1296 0.0332 0.2289 0.0050 0.0097 0.5935 Precision/Correctness 0.943 0.978 0.671 0.947 0.922 0.725 Recall/Completeness 0.935 0.941 0.917 0.768 0.972 0.594 F1 0.939 0.959 0.775 0.848 0.947 0.653 Overall accuracy 0.889 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 0.9196 0.0181 0.0175 0.0217 0.0021 0.0211 building 0.0189 0.9280 0.0051 0.0019 0.0004 0.0457 low_veg 0.0464 0.0097 0.8033 0.1325 0.0002 0.0079 tree 0.0360 0.0097 0.1372 0.8150 0.0006 0.0015 car 0.0441 0.0224 0.0008 0.0309 0.8826 0.0192 clutter 0.1738 0.0143 0.1041 0.0020 0.0076 0.6982 Precision/Correctness 0.942 0.944 0.731 0.716 0.866 0.743 Recall/Completeness 0.920 0.928 0.803 0.815 0.883 0.698 F1 0.931 0.936 0.765 0.762 0.874 0.720 Overall accuracy 0.881 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 0.9344 0.0150 0.0147 0.0152 0.0005 0.0203 building 0.0161 0.9336 0.0040 0.0012 0.0004 0.0447 low_veg 0.0340 0.0063 0.8275 0.1251 0.0001 0.0071 tree 0.0263 0.0078 0.1293 0.8348 0.0004 0.0013 car 0.0044 0.0290 0.0001 0.0208 0.9274 0.0182 clutter 0.1662 0.0129 0.1027 0.0014 0.0069 0.7098 Precision/Correctness 0.951 0.955 0.745 0.751 0.910 0.753 Recall/Completeness 0.934 0.934 0.828 0.835 0.927 0.710 F1 0.942 0.944 0.784 0.791 0.918 0.731 Overall accuracy 0.896 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image