VectorOutputConfig#

Note

All Configs are derived from rastervision.pipeline.config.Config, which itself is a pydantic Model.

pydantic model VectorOutputConfig[source]#

Config for vectorized semantic segmentation predictions.

Show JSON schema

{
   "title": "VectorOutputConfig",
   "description": "Config for vectorized semantic segmentation predictions.",
   "type": "object",
   "properties": {
      "class_id": {
         "description": "The prediction class that is to turned into vectors.",
         "title": "Class Id",
         "type": "integer"
      },
      "denoise": {
         "default": 8,
         "description": "Diameter of the circular structural element used to remove high-frequency signals from the image. Smaller values will reduce less noise and make vectorization slower and more memory intensive (especially for large images). Larger values will remove more noise and make vectorization faster but might also remove legitimate detections.",
         "title": "Denoise",
         "type": "integer"
      },
      "threshold": {
         "anyOf": [
            {
               "type": "number"
            },
            {
               "type": "null"
            }
         ],
         "default": null,
         "description": "Probability threshold for creating the binary mask for the pixels of this class. Pixels will be considered to belong to this class if their probability for this class is >= ``threshold``. Note that Raster Vision treats classes as mutually exclusive so the threshold should vary with the number of total classes. ``None`` is equivalent to setting this to (1 / num_classes). Defaults to ``None``.",
         "title": "Threshold"
      },
      "type_hint": {
         "const": "vector_output",
         "default": "vector_output",
         "enum": [
            "vector_output"
         ],
         "title": "Type Hint",
         "type": "string"
      }
   },
   "additionalProperties": false,
   "required": [
      "class_id"
   ]
}

Config:

extra: str = forbid
validate_assignment: bool = True

Fields:

class_id (int)
denoise (int)
threshold (float | None)
type_hint (Literal['vector_output'])

field class_id: int [Required]#: The prediction class that is to turned into vectors.

field denoise: int = 8#: Diameter of the circular structural element used to remove high-frequency signals from the image. Smaller values will reduce less noise and make vectorization slower and more memory intensive (especially for large images). Larger values will remove more noise and make vectorization faster but might also remove legitimate detections.

field threshold: float | None = None#: Probability threshold for creating the binary mask for the pixels of this class. Pixels will be considered to belong to this class if their probability for this class is >= threshold. Note that Raster Vision treats classes as mutually exclusive so the threshold should vary with the number of total classes. None is equivalent to setting this to (1 / num_classes). Defaults to None.

field type_hint: Literal['vector_output'] = 'vector_output'#

build()#

Build an instance of the corresponding type of object using this config.

For example, BackendConfig will build a Backend object. The arguments to this method will vary depending on the type of Config.

classmethod deserialize(inp: str | dict | Config) → Self#

Deserialize Config from a JSON file or dict, upgrading if possible.

If inp is already a Config, it is returned as is.

Parameters:: inp (str | dict | Config) – a URI to a JSON file or a dict.
Return type:: Self

classmethod from_dict(cfg_dict: dict) → Self#

Deserialize Config from a dict.

Parameters:: cfg_dict (dict) – Dict to deserialize.
Return type:: Self

classmethod from_file(uri: str) → Self#

Deserialize Config from a JSON file, upgrading if possible.

Parameters:: uri (str) – URI to load from.
Return type:: Self

get_uri(root: str, class_config: ClassConfig | None = None) → str[source]#

Parameters:

root (str) –
class_config (ClassConfig | None) –

Return type:

str

recursive_validate_config()#

Recursively validate hierarchies of Configs.

This uses reflection to call validate_config on a hierarchy of Configs using a depth-first pre-order traversal.

revalidate()#

Re-validate an instantiated Config.

Runs all Pydantic validators plus self.validate_config().

to_file(uri: str, with_rv_metadata: bool = True) → None#

Save a Config to a JSON file, optionally with RV metadata.

Parameters:

uri (str) – URI to save to.
with_rv_metadata (bool) – If True, inject Raster Vision metadata such as plugin_versions, so that the config can be upgraded when loaded.

Return type:

None

update(*args, **kwargs)#

Update any fields before validation.

Subclasses should override this to provide complex default behavior, for example, setting default values as a function of the values of other fields. The arguments to this method will vary depending on the type of Config.

validate_config()#

Validate fields that should be checked after update is called.

This is to complement the builtin validation that Pydantic performs at the time of object construction.

validate_list(field: str, valid_options: list[str])#

Validate a list field.

Parameters:

field (str) – name of field to validate
valid_options (list[str]) – values that field is allowed to take

Raises:

ConfigError – if field is invalid

vectorize(mask: np.ndarray) → Iterator[BaseGeometry][source]#

Vectorize binary mask representing the target class into polygons.

Parameters:: mask (np.ndarray) –
Return type:: Iterator[BaseGeometry]