Facility Extraction API

Module: mls.facility. This page documents 1 public functions with exact tool IDs, complete parameters, Parameters fields, and examples.

Back to the Script Tool API Reference

Functions

  • road_facility_detection — Detect and vectorize road facilities (traffic signs, traffic lights, manhole covers, etc.) from panoramic/planar images using deep learning.

road_facility_detection

Detect and vectorize road facilities (traffic signs, traffic lights, manhole covers, etc.) from panoramic/planar images using deep learning.

  • Exact tool ID:Facility_Road_Facility_Detection
  • Recommended call:mls.facility.road_facility_detection(...)
  • Alias:Road_Facility_Detection
  • Required parameters:input_paths, Parameters

Parameters

Parameter Type Required Default Description
input_paths array Yes — Camera file paths (panoramic or planar) for facility detection.
Parameters object Yes — Facility detection parameters

Parameters fields

Field Type Required Default Description
fromClass object No — Class selection. JSON object mapping selection index to classification code. Example: {"0":2}
type integer No 0 Run mode: 0 = Built-in model, 2 = Custom mode
GPU boolean No True Whether to use GPU mode
GPUID integer No 0 GPU device ID
cameraType integer No 0 Camera type: 0 = panoramic, 1 = planar
startPic integer No 0 Start image index
endPic integer No 999999999 End image index
panoramaCameraFile string No — Panoramic camera file path
planarCameraFile string No — Planar camera file path
planarCameraName string No — Planar camera name
minDis number No 2.0 Minimum distance for object filtering
maxDis number No 20.0 Maximum distance for object filtering
batchCell number No 50.0 Batch processing grid size
warningConf number No 0.5 Warning sign confidence threshold
injunctionConf number No 0.5 Injunction sign confidence threshold
instructionConf number No 0.5 Instruction sign confidence threshold
guideConf number No 0.5 Guide sign confidence threshold
travelConf number No 0.5 Travel sign confidence threshold
limitConf number No 0.5 Limit sign confidence threshold
trafficLightConf number No 0.6 Traffic light confidence threshold
auxiliaryConf number No 0.5 Auxiliary facility confidence threshold
manholeConf number No 0.7 Manhole cover confidence threshold
sewageGrateConf number No 0.7 Sewage grate confidence threshold
dustbinConf number No 0.3 Dustbin confidence threshold
monitorConf number No 0.3 Monitor confidence threshold
streetLightConf number No 0.7 Street light confidence threshold
busConf number No 0.3 Bus station confidence threshold
roadConstruction number No 0.5 Road construction confidence threshold
transformer number No 0.3 Transformer confidence threshold
pole number No 1.1 Pole confidence threshold
customRunType integer No -1 Custom mode type: 0 = model mode, 1 = json mode, -1 = invalid
customModelFullPath string No — Custom model full path
jsonDirFullPath string No — JSON directory full path
distRadius number No 0 Horizontal distance radius for camera point cloud range
targetPointLayerID string No — Target point layer ID for custom mode

Example

from gvscript import mls

params = mls.facility.road_facility_detection_parameters()
params.fromClass = {}

result = mls.facility.road_facility_detection(
    Parameters=params,
    input_paths=['D:/data/input.LiData'],
)
print(result.ok, result.output, result.message)

results matching ""

    No results matching ""