graph TD
In[Input Polygon] --> Decon[Deconstruct: Exterior & Interiors]
Decon --> Buff[Buffer Input to Find Network]
Buff --> Align[Align Segments to Network Elements]
Align --> Recon[Reconstruct Polygon Rings]
Recon --> Post[Post-processing & Sliver Removal]
Post --> End[Final ProcessResult]
NetworkGeometryProcessor
processor.NetworkGeometryProcessor(config, feedback=None)Processor that aligns geometries based on a linear network.
This processor decomposes polygons into their exterior and interior linear rings and aligns these boundaries to the linear elements (lines and points) found in the reference dataset.
Attributes
| Name | Type | Description |
|---|---|---|
| processor_id | ProcessorID | The unique identifier for this processor (ProcessorID.NETWORK). |
Methods
| Name | Description |
|---|---|
| process | Aligns the boundaries of the input geometry to a reference network. |
process
processor.NetworkGeometryProcessor.process(
input_geometry,
reference_data,
mitre_limit,
correction_distance,
relevant_distance,
**kwargs,
)Aligns the boundaries of the input geometry to a reference network.
The process buffers the input to find relevant network elements, processes exterior and interior rings separately, and reconstructs the polygon after alignment.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
| input_geometry | BaseGeometry | The thematic geometry (Polygon or MultiPolygon) to align. | required |
| reference_data | AlignerFeatureCollection | The reference dataset, specifically using its elements property. |
required |
| mitre_limit | float | Mitre limit for buffering operations. | required |
| correction_distance | float | Distance used for cleaning and noise reduction. | required |
| relevant_distance | float | The maximum distance to search for network elements. | required |
| **kwargs | Any | Additional arguments passed to the processor. | {} |
Returns
| Name | Type | Description |
|---|---|---|
| ProcessResult | A dictionary containing the reconstructed and cleaned polygon. |
Notes
The network processing follows a “deconstruct-align-reconstruct” flow: