- accepts: idc.api.ObjectDetectionData
Saves the bounding box/polygon definitions in YOLO .txt format. By default, places images in the 'images' subdir and the annotations in 'labels'.
usage: to-yolo-od [-h] [-l {DEBUG,INFO,WARNING,ERROR,CRITICAL}]
[-N LOGGER_NAME] [-r SPLIT_RATIOS [SPLIT_RATIOS ...]]
[-n SPLIT_NAMES [SPLIT_NAMES ...]] -o OUTPUT
[--image_subdir DIR] [--labels_subdir DIR] [-p]
[--categories [CATEGORIES [CATEGORIES ...]]]
[--labels LABELS] [--labels_csv LABELS_CSV]
[--annotations_only]
Saves the bounding box/polygon definitions in YOLO .txt format. By default,
places images in the 'images' subdir and the annotations in 'labels'.
optional arguments:
-h, --help show this help message and exit
-l {DEBUG,INFO,WARNING,ERROR,CRITICAL}, --logging_level {DEBUG,INFO,WARNING,ERROR,CRITICAL}
The logging level to use. (default: WARN)
-N LOGGER_NAME, --logger_name LOGGER_NAME
The custom name to use for the logger, uses the plugin
name by default (default: None)
-r SPLIT_RATIOS [SPLIT_RATIOS ...], --split_ratios SPLIT_RATIOS [SPLIT_RATIOS ...]
The split ratios to use for generating the splits
(must sum up to 100) (default: None)
-n SPLIT_NAMES [SPLIT_NAMES ...], --split_names SPLIT_NAMES [SPLIT_NAMES ...]
The split names to use for the generated splits.
(default: None)
-o OUTPUT, --output OUTPUT
The directory to store the images/.txt files in. Any
defined splits get added beneath there. (default:
None)
--image_subdir DIR The name of the sub-dir to use for storing the images
in. (default: None)
--labels_subdir DIR The name of the sub-dir to use for storing the
annotations in. (default: None)
-p, --use_polygon_format
Whether to write the annotations in polygon format
rather than bbox format (default: False)
--categories [CATEGORIES [CATEGORIES ...]]
The predefined order of categories. (default: None)
--labels LABELS The text file (no path) with the comma-separated list
of labels (default: None)
--labels_csv LABELS_CSV
The CSV file (no path) to write the label mapping to
(index and label) (default: None)
--annotations_only Outputs only the annotations and skips the base image.
(default: False)