labelme data_annotated --labels labels.txt --nodata --validatelabel exact --config '{shift_auto_shape_color: -2}'
# It generates:
# - data_dataset_voc/JPEGImages
# - data_dataset_voc/SegmentationClass
# - data_dataset_voc/SegmentationClassVisualization
./labelme2voc.py data_annotated data_dataset_voc --labels labels.txt
Fig 1. JPEG image (left), PNG label (center), JPEG label visualization (right)
Note that the label file contains only very low label values (ex. 0, 4, 14
), and
255
indicates the __ignore__
label value (-1
in the npy file).
You can see the label PNG file by following.
labelme_draw_label_png data_dataset_voc/SegmentationClassPNG/2011_000003.png