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 0.912 0.020 0.037 0.018 0.003 0.011 building 0.023 0.956 0.010 0.003 0.000 0.007 low_veg 0.043 0.010 0.876 0.057 0.000 0.013 tree 0.033 0.005 0.132 0.824 0.002 0.004 car 0.082 0.004 0.003 0.017 0.882 0.011 clutter 0.286 0.114 0.182 0.021 0.010 0.387 Precision/Correctness 0.892 0.940 0.804 0.877 0.907 0.668 Recall/Completeness 0.912 0.956 0.876 0.824 0.882 0.387 F1 0.902 0.948 0.838 0.850 0.895 0.490 Overall accuracy 0.875	Overall statistics, reference set: no_boundary Number of non processed tiles by this participant: 0 ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.932 0.016 0.030 0.013 0.000 0.009 building 0.019 0.965 0.007 0.002 0.000 0.007 low_veg 0.032 0.006 0.902 0.048 0.000 0.012 tree 0.024 0.004 0.117 0.850 0.002 0.003 car 0.016 0.004 0.001 0.012 0.955 0.012 clutter 0.282 0.100 0.172 0.017 0.011 0.420 Precision/Correctness 0.914 0.956 0.829 0.904 0.947 0.694 Recall/Completeness 0.932 0.965 0.902 0.850 0.955 0.420 F1 0.923 0.960 0.864 0.876 0.951 0.523 Overall accuracy 0.899
============================================================= Tile top_potsdam_2_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8655 0.0381 0.0543 0.0155 0.0027 0.0238 building 0.0106 0.9537 0.0198 0.0052 0.0001 0.0106 low_veg 0.0285 0.0126 0.8862 0.0411 0.0002 0.0313 tree 0.0291 0.0115 0.1373 0.8146 0.0011 0.0063 car 0.0815 0.0087 0.0032 0.0123 0.8587 0.0355 clutter 0.1260 0.1953 0.1395 0.0328 0.0087 0.4977 Precision/Correctness 0.911 0.899 0.839 0.913 0.895 0.333 Recall/Completeness 0.865 0.954 0.886 0.815 0.859 0.498 F1 0.888 0.926 0.862 0.861 0.876 0.399 Overall accuracy 0.869 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 0.8929 0.0327 0.0438 0.0096 0.0004 0.0206 building 0.0073 0.9632 0.0168 0.0039 0.0000 0.0088 low_veg 0.0220 0.0079 0.9098 0.0294 0.0001 0.0308 tree 0.0214 0.0097 0.1173 0.8455 0.0007 0.0054 car 0.0220 0.0094 0.0020 0.0066 0.9176 0.0424 clutter 0.0946 0.1856 0.1068 0.0235 0.0098 0.5797 Precision/Correctness 0.937 0.923 0.865 0.941 0.951 0.318 Recall/Completeness 0.893 0.963 0.910 0.845 0.918 0.580 F1 0.915 0.943 0.887 0.891 0.934 0.411 Overall accuracy 0.896 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 0.8749 0.0142 0.0841 0.0146 0.0018 0.0104 building 0.0050 0.9774 0.0060 0.0056 0.0001 0.0059 low_veg 0.0134 0.0046 0.9464 0.0292 0.0000 0.0064 tree 0.0132 0.0036 0.1972 0.7821 0.0008 0.0031 car 0.0646 0.0005 0.0044 0.0141 0.9141 0.0022 clutter 0.1023 0.2449 0.1496 0.0622 0.0194 0.4215 Precision/Correctness 0.922 0.915 0.823 0.949 0.896 0.208 Recall/Completeness 0.875 0.977 0.946 0.782 0.914 0.422 F1 0.898 0.945 0.880 0.857 0.905 0.279 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 0.9100 0.0108 0.0612 0.0087 0.0001 0.0091 building 0.0029 0.9874 0.0024 0.0033 0.0001 0.0039 low_veg 0.0094 0.0025 0.9601 0.0223 0.0000 0.0057 tree 0.0096 0.0030 0.1806 0.8036 0.0005 0.0028 car 0.0104 0.0002 0.0026 0.0085 0.9758 0.0025 clutter 0.0772 0.2761 0.1027 0.0643 0.0208 0.4590 Precision/Correctness 0.945 0.940 0.842 0.964 0.946 0.190 Recall/Completeness 0.910 0.987 0.960 0.804 0.976 0.459 F1 0.927 0.963 0.897 0.876 0.961 0.269 Overall accuracy 0.897 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 0.8882 0.0195 0.0557 0.0203 0.0036 0.0127 building 0.0149 0.9643 0.0092 0.0042 0.0000 0.0073 low_veg 0.0540 0.0200 0.8584 0.0538 0.0002 0.0134 tree 0.0271 0.0068 0.1034 0.8561 0.0019 0.0046 car 0.0813 0.0041 0.0029 0.0216 0.8804 0.0097 clutter 0.2184 0.1408 0.2754 0.0196 0.0032 0.3426 Precision/Correctness 0.885 0.943 0.809 0.911 0.911 0.466 Recall/Completeness 0.888 0.964 0.858 0.856 0.880 0.343 F1 0.887 0.953 0.833 0.883 0.896 0.395 Overall accuracy 0.880 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 0.9171 0.0133 0.0450 0.0132 0.0003 0.0111 building 0.0110 0.9737 0.0062 0.0030 0.0000 0.0061 low_veg 0.0435 0.0112 0.8910 0.0434 0.0001 0.0108 tree 0.0197 0.0054 0.0857 0.8839 0.0014 0.0039 car 0.0147 0.0047 0.0016 0.0148 0.9543 0.0099 clutter 0.2026 0.1370 0.2531 0.0121 0.0036 0.3916 Precision/Correctness 0.915 0.963 0.841 0.936 0.964 0.480 Recall/Completeness 0.917 0.974 0.891 0.884 0.954 0.392 F1 0.916 0.968 0.866 0.909 0.959 0.431 Overall accuracy 0.909 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 0.8631 0.0249 0.0726 0.0247 0.0033 0.0115 building 0.0107 0.9761 0.0052 0.0029 0.0000 0.0050 low_veg 0.0244 0.0124 0.9123 0.0423 0.0000 0.0086 tree 0.0191 0.0035 0.1978 0.7725 0.0013 0.0059 car 0.0799 0.0079 0.0022 0.0331 0.8709 0.0060 clutter 0.1877 0.1683 0.2125 0.0267 0.0059 0.3989 Precision/Correctness 0.898 0.936 0.770 0.915 0.895 0.619 Recall/Completeness 0.863 0.976 0.912 0.772 0.871 0.399 F1 0.880 0.956 0.835 0.838 0.883 0.485 Overall accuracy 0.864 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_3_14_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.8958 0.0189 0.0583 0.0170 0.0003 0.0098 building 0.0081 0.9838 0.0022 0.0019 0.0000 0.0040 low_veg 0.0157 0.0072 0.9382 0.0323 0.0000 0.0066 tree 0.0143 0.0025 0.1812 0.7959 0.0011 0.0050 car 0.0172 0.0080 0.0013 0.0229 0.9440 0.0066 clutter 0.1746 0.1596 0.1808 0.0209 0.0062 0.4579 Precision/Correctness 0.928 0.955 0.794 0.940 0.945 0.658 Recall/Completeness 0.896 0.984 0.938 0.796 0.944 0.458 F1 0.911 0.969 0.860 0.862 0.945 0.540 Overall accuracy 0.890 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 0.9259 0.0076 0.0335 0.0247 0.0021 0.0063 building 0.0239 0.9573 0.0081 0.0048 0.0003 0.0056 low_veg 0.0514 0.0068 0.8453 0.0911 0.0001 0.0052 tree 0.0255 0.0022 0.0535 0.9157 0.0014 0.0016 car 0.0873 0.0018 0.0032 0.0304 0.8742 0.0031 clutter 0.4801 0.0998 0.1613 0.0635 0.0115 0.1837 Precision/Correctness 0.841 0.968 0.800 0.811 0.918 0.723 Recall/Completeness 0.926 0.957 0.845 0.916 0.874 0.184 F1 0.881 0.963 0.822 0.860 0.896 0.293 Overall accuracy 0.872 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 0.9464 0.0040 0.0263 0.0178 0.0003 0.0052 building 0.0173 0.9701 0.0039 0.0035 0.0003 0.0049 low_veg 0.0390 0.0023 0.8748 0.0797 0.0000 0.0041 tree 0.0147 0.0011 0.0395 0.9428 0.0009 0.0011 car 0.0134 0.0016 0.0006 0.0209 0.9607 0.0028 clutter 0.4981 0.0860 0.1445 0.0548 0.0131 0.2036 Precision/Correctness 0.868 0.980 0.840 0.849 0.939 0.752 Recall/Completeness 0.946 0.970 0.875 0.943 0.961 0.204 F1 0.905 0.975 0.857 0.893 0.950 0.320 Overall accuracy 0.899 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 0.8817 0.0262 0.0633 0.0127 0.0040 0.0121 building 0.0225 0.9555 0.0071 0.0065 0.0002 0.0083 low_veg 0.0785 0.0140 0.8270 0.0525 0.0002 0.0277 tree 0.0338 0.0045 0.1385 0.8116 0.0023 0.0093 car 0.0713 0.0041 0.0077 0.0218 0.8754 0.0198 clutter 0.2690 0.0571 0.1973 0.0101 0.0003 0.4661 Precision/Correctness 0.797 0.921 0.750 0.894 0.926 0.776 Recall/Completeness 0.882 0.956 0.827 0.812 0.875 0.466 F1 0.837 0.938 0.787 0.851 0.900 0.582 Overall accuracy 0.829 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 0.9021 0.0225 0.0563 0.0081 0.0004 0.0105 building 0.0184 0.9645 0.0047 0.0051 0.0002 0.0071 low_veg 0.0659 0.0088 0.8538 0.0442 0.0001 0.0272 tree 0.0252 0.0032 0.1195 0.8417 0.0018 0.0086 car 0.0169 0.0041 0.0065 0.0156 0.9362 0.0208 clutter 0.2701 0.0538 0.1849 0.0074 0.0001 0.4836 Precision/Correctness 0.817 0.937 0.772 0.919 0.971 0.799 Recall/Completeness 0.902 0.965 0.854 0.842 0.936 0.484 F1 0.857 0.950 0.811 0.879 0.953 0.602 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_15_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9081 0.0293 0.0472 0.0054 0.0028 0.0072 building 0.0254 0.9515 0.0111 0.0021 0.0001 0.0099 low_veg 0.0530 0.0164 0.8888 0.0303 0.0005 0.0111 tree 0.0614 0.0080 0.1428 0.7774 0.0058 0.0046 car 0.0930 0.0134 0.0026 0.0084 0.8718 0.0108 clutter 0.2674 0.2304 0.2011 0.0144 0.0075 0.2793 Precision/Correctness 0.900 0.935 0.769 0.925 0.911 0.497 Recall/Completeness 0.908 0.951 0.889 0.777 0.872 0.279 F1 0.904 0.943 0.825 0.845 0.891 0.358 Overall accuracy 0.883 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image	============================================================= Tile top_potsdam_4_15_class.tif, reference set: no_boundary ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9271 0.0234 0.0405 0.0031 0.0003 0.0056 building 0.0222 0.9586 0.0088 0.0014 0.0001 0.0089 low_veg 0.0337 0.0085 0.9252 0.0240 0.0001 0.0085 tree 0.0447 0.0063 0.1208 0.8203 0.0042 0.0037 car 0.0189 0.0156 0.0006 0.0041 0.9481 0.0126 clutter 0.2600 0.2297 0.1977 0.0102 0.0070 0.2954 Precision/Correctness 0.928 0.951 0.797 0.949 0.954 0.507 Recall/Completeness 0.927 0.959 0.925 0.820 0.948 0.295 F1 0.927 0.955 0.856 0.880 0.951 0.373 Overall accuracy 0.909 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 0.9240 0.0167 0.0274 0.0230 0.0022 0.0067 building 0.0194 0.9653 0.0063 0.0036 0.0000 0.0054 low_veg 0.0647 0.0123 0.8311 0.0847 0.0001 0.0072 tree 0.0466 0.0049 0.1008 0.8443 0.0021 0.0013 car 0.0975 0.0015 0.0005 0.0080 0.8830 0.0094 clutter 0.2692 0.2437 0.1611 0.0159 0.0510 0.2591 Precision/Correctness 0.898 0.944 0.848 0.803 0.896 0.599 Recall/Completeness 0.924 0.965 0.831 0.844 0.883 0.259 F1 0.911 0.954 0.839 0.823 0.889 0.362 Overall accuracy 0.885 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 0.9432 0.0114 0.0228 0.0172 0.0002 0.0053 building 0.0151 0.9720 0.0050 0.0029 0.0000 0.0050 low_veg 0.0568 0.0071 0.8552 0.0752 0.0000 0.0057 tree 0.0320 0.0032 0.0878 0.8745 0.0016 0.0010 car 0.0160 0.0014 0.0000 0.0049 0.9679 0.0098 clutter 0.2419 0.2314 0.1537 0.0139 0.0649 0.2942 Precision/Correctness 0.925 0.962 0.874 0.832 0.916 0.612 Recall/Completeness 0.943 0.972 0.855 0.874 0.968 0.294 F1 0.934 0.967 0.865 0.853 0.941 0.397 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 0.9212 0.0125 0.0246 0.0266 0.0017 0.0134 building 0.0131 0.9762 0.0058 0.0016 0.0000 0.0032 low_veg 0.0654 0.0095 0.8298 0.0825 0.0003 0.0124 tree 0.0478 0.0048 0.0889 0.8528 0.0030 0.0026 car 0.1021 0.0016 0.0057 0.0176 0.8474 0.0254 clutter 0.2206 0.1838 0.1776 0.0325 0.0097 0.3758 Precision/Correctness 0.850 0.925 0.823 0.845 0.904 0.778 Recall/Completeness 0.921 0.976 0.830 0.853 0.847 0.376 F1 0.884 0.950 0.826 0.849 0.875 0.507 Overall accuracy 0.862 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 0.9404 0.0098 0.0171 0.0200 0.0001 0.0126 building 0.0104 0.9822 0.0036 0.0010 0.0000 0.0029 low_veg 0.0509 0.0064 0.8544 0.0769 0.0000 0.0114 tree 0.0374 0.0039 0.0791 0.8751 0.0022 0.0023 car 0.0247 0.0010 0.0044 0.0112 0.9302 0.0286 clutter 0.2037 0.1599 0.1697 0.0297 0.0103 0.4267 Precision/Correctness 0.881 0.944 0.849 0.869 0.929 0.798 Recall/Completeness 0.940 0.982 0.854 0.875 0.930 0.427 F1 0.910 0.963 0.851 0.872 0.930 0.556 Overall accuracy 0.887 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 0.8997 0.0286 0.0420 0.0190 0.0019 0.0089 building 0.0209 0.9580 0.0101 0.0052 0.0006 0.0052 low_veg 0.0451 0.0060 0.8396 0.1011 0.0001 0.0080 tree 0.0324 0.0057 0.0798 0.8772 0.0018 0.0032 car 0.1158 0.0015 0.0023 0.0147 0.8547 0.0110 clutter 0.2980 0.1315 0.1720 0.0131 0.0045 0.3810 Precision/Correctness 0.901 0.914 0.747 0.805 0.940 0.800 Recall/Completeness 0.900 0.958 0.840 0.877 0.855 0.381 F1 0.901 0.936 0.791 0.840 0.895 0.516 Overall accuracy 0.865 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 0.9148 0.0253 0.0368 0.0143 0.0005 0.0082 building 0.0155 0.9671 0.0084 0.0039 0.0005 0.0046 low_veg 0.0284 0.0021 0.8681 0.0941 0.0000 0.0072 tree 0.0226 0.0040 0.0687 0.9005 0.0013 0.0029 car 0.0264 0.0011 0.0006 0.0124 0.9471 0.0124 clutter 0.2943 0.1002 0.1675 0.0107 0.0042 0.4230 Precision/Correctness 0.924 0.934 0.769 0.835 0.963 0.820 Recall/Completeness 0.915 0.967 0.868 0.901 0.947 0.423 F1 0.919 0.950 0.815 0.867 0.955 0.558 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_13_class.tif, reference set: full_reference ↓ predicted || reference → imp_surf building low_veg tree car clutter imp_surf 0.9273 0.0192 0.0234 0.0176 0.0018 0.0107 building 0.0358 0.9382 0.0175 0.0011 0.0003 0.0072 low_veg 0.0194 0.0047 0.9356 0.0386 0.0000 0.0018 tree 0.0592 0.0026 0.0754 0.8605 0.0012 0.0011 car 0.0668 0.0019 0.0003 0.0053 0.9222 0.0034 clutter 0.3276 0.0326 0.3247 0.0288 0.0067 0.2796 Precision/Correctness 0.926 0.960 0.885 0.860 0.917 0.474 Recall/Completeness 0.927 0.938 0.936 0.860 0.922 0.280 F1 0.927 0.949 0.910 0.860 0.920 0.352 Overall accuracy 0.911 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 0.9445 0.0167 0.0179 0.0114 0.0002 0.0093 building 0.0307 0.9485 0.0136 0.0006 0.0003 0.0065 low_veg 0.0111 0.0021 0.9533 0.0323 0.0000 0.0012 tree 0.0426 0.0015 0.0630 0.8912 0.0009 0.0008 car 0.0105 0.0018 0.0002 0.0027 0.9816 0.0033 clutter 0.3192 0.0297 0.3336 0.0268 0.0063 0.2845 Precision/Correctness 0.944 0.969 0.905 0.891 0.964 0.484 Recall/Completeness 0.944 0.948 0.953 0.891 0.982 0.284 F1 0.944 0.959 0.929 0.891 0.973 0.358 Overall accuracy 0.930 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 0.9516 0.0097 0.0211 0.0105 0.0042 0.0029 building 0.0400 0.9463 0.0056 0.0014 0.0002 0.0065 low_veg 0.0513 0.0118 0.8666 0.0622 0.0005 0.0076 tree 0.0451 0.0069 0.0785 0.8631 0.0052 0.0013 car 0.0639 0.0036 0.0016 0.0131 0.9147 0.0031 clutter 0.2563 0.2419 0.2321 0.0201 0.0167 0.2329 Precision/Correctness 0.920 0.960 0.854 0.845 0.915 0.545 Recall/Completeness 0.952 0.946 0.867 0.863 0.915 0.233 F1 0.936 0.953 0.860 0.854 0.915 0.326 Overall accuracy 0.910 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 0.9680 0.0063 0.0156 0.0074 0.0003 0.0024 building 0.0342 0.9558 0.0029 0.0008 0.0001 0.0061 low_veg 0.0342 0.0067 0.8975 0.0554 0.0001 0.0061 tree 0.0341 0.0056 0.0654 0.8896 0.0041 0.0012 car 0.0069 0.0024 0.0004 0.0090 0.9779 0.0034 clutter 0.2295 0.2712 0.2261 0.0163 0.0161 0.2408 Precision/Correctness 0.941 0.969 0.882 0.875 0.966 0.558 Recall/Completeness 0.968 0.956 0.898 0.890 0.978 0.241 F1 0.954 0.962 0.890 0.882 0.972 0.336 Overall accuracy 0.931 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 0.9161 0.0205 0.0305 0.0091 0.0048 0.0190 building 0.0220 0.9469 0.0222 0.0033 0.0000 0.0056 low_veg 0.0451 0.0058 0.8650 0.0469 0.0010 0.0362 tree 0.0464 0.0041 0.1548 0.7894 0.0034 0.0020 car 0.0672 0.0035 0.0049 0.0166 0.9005 0.0073 clutter 0.1452 0.0540 0.2317 0.0062 0.0109 0.5520 Precision/Correctness 0.921 0.946 0.701 0.911 0.897 0.697 Recall/Completeness 0.916 0.947 0.865 0.789 0.901 0.552 F1 0.918 0.946 0.775 0.846 0.899 0.616 Overall accuracy 0.874 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 0.9354 0.0181 0.0220 0.0058 0.0005 0.0181 building 0.0160 0.9575 0.0192 0.0023 0.0000 0.0049 low_veg 0.0282 0.0034 0.8940 0.0381 0.0001 0.0361 tree 0.0366 0.0029 0.1370 0.8193 0.0026 0.0016 car 0.0111 0.0032 0.0012 0.0120 0.9646 0.0080 clutter 0.1354 0.0537 0.2254 0.0049 0.0109 0.5697 Precision/Correctness 0.942 0.954 0.724 0.936 0.950 0.714 Recall/Completeness 0.935 0.958 0.894 0.819 0.965 0.570 F1 0.939 0.956 0.800 0.874 0.957 0.634 Overall accuracy 0.896 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 0.9384 0.0132 0.0110 0.0271 0.0021 0.0083 building 0.0415 0.9345 0.0018 0.0020 0.0001 0.0201 low_veg 0.0603 0.0075 0.7728 0.1502 0.0002 0.0091 tree 0.0230 0.0067 0.1109 0.8566 0.0004 0.0024 car 0.0544 0.0063 0.0003 0.0326 0.8975 0.0090 clutter 0.3962 0.0113 0.0939 0.0030 0.0136 0.4820 Precision/Correctness 0.897 0.959 0.773 0.691 0.841 0.816 Recall/Completeness 0.938 0.935 0.773 0.857 0.898 0.482 F1 0.917 0.947 0.773 0.765 0.868 0.606 Overall accuracy 0.873 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 0.9538 0.0100 0.0083 0.0197 0.0004 0.0078 building 0.0377 0.9407 0.0009 0.0012 0.0001 0.0194 low_veg 0.0469 0.0036 0.7997 0.1418 0.0000 0.0080 tree 0.0156 0.0050 0.1043 0.8726 0.0002 0.0022 car 0.0082 0.0077 0.0000 0.0239 0.9508 0.0093 clutter 0.3918 0.0101 0.0931 0.0024 0.0127 0.4898 Precision/Correctness 0.904 0.970 0.789 0.725 0.879 0.828 Recall/Completeness 0.954 0.941 0.800 0.873 0.951 0.490 F1 0.928 0.955 0.794 0.792 0.914 0.615 Overall accuracy 0.887 Red/green image, indicating wrongly classified pixels Original, resized true ortho image for better comparison Classified image