kingsley0107 / road_regularization

🚀 Road network simplification, batch conversion of OSM road multi-lines to single lines 🚀 Detection and export of dead-end roads 🚀 Parcel division 🚀 Road network optimization 🚀 Block divide, cartography, export dead-end roads, and so on...

English 中文版

road_geo_regularization

A universal simplification of the main road network in the city, which can be used for non-high precision land division, etc.

Performance:

raw road data

• 

simplified road data

• 

Environment:

• Arcgis Pro
• Python 3.x from Arcgis Pro
• Arcpy from Arcgis Pro
• Encoding utf-8_sig

Required Data:

• Road network data
• Boundary data
• Data needs to be placed in the path specified in config.py in advance, pay attention to the naming format

User Manual:

1. Specify directories manually in config/config.py

   • Configuration options:
     • ./config/config.py needs to be manually specified:
     • GDB file path (automatically created if it doesn't exist), GDB will be automatically cleared if all steps are executed successfully
     • Road network data root directory (pay attention to the naming conventions of road network data)
     • Boundary data root directory (pay attention to the naming conventions)
   • Road types:
     • Currently, road types are selected mainly for the purpose of parcel division. The road network categories involved in the simplification are filtered based on the ROAD_TYPE_FIELD in config.py. You can add a new category in config.py for customized fine road network processing.

2. Execute main.py

   • The main function start_process in main.py accepts the following parameters:
     • CITY: City name, ensure the naming is the same as the suffix of the road network (see comments for details)
     • MODE: Simplification mode, must be 'mixed'
     • smooth_level: Level of simplification, the higher the value, the smoother the result (distortion)
     • extend_distance: Distance of extending dead-end roads, the higher the value, the more closed road networks (but may extend a road where it does not exist)
     • spike_keep: Threshold for cleaning small spikes, roads with a length lower than this threshold will be removed, the higher the value, the more regular the roads, but it may remove roads that are actually dead-ends

Road Network Simplification Principle:

1. Extract CenterLines

   • 

     load raw data.

   • 

     build buffer for each line.

   • 

     dissolve the buffers.

   • 

     Extract centerlines from merged buffers

   • 

   Extraction result.

2. road optimization:

   • 

     Extension of non-closed roads (in blue).

   • 

     Extraction of dead-end roads (points).

   • 

     Filtering small spikes (in red) with a length below the threshold.

3. Road Network Spatial Information Re-association:：

   • 

   Create buffers based on the simplified road network and load the original road network data for information assignment.

   • 

   For each buffer, the matching rule with the original road network is to select the most representative road within the buffer (largest overlap).

   • 

     eg. The most representative road within this buffer is enclosed.

4. output result

   • 

A Hidden Parameter:

There is a hidden parameter called keep_spike in the clean_spike function in simplify.py, which is not commonly used.

It may be useful when plotting data related to dead-end roads. If needed, you can manually set keep_spike to True.

The data for dead-end points and dead-end lines will be output to the result path, named 'err_points.shp' and 'err_lines.shp', respectively.