Workflow on Automated Pole Detector

This page describes the workflow of feature extraction using the automated pole detector.
Overview and concepts about automated feature extraction, see Automated Feature Extraction.
This tool is available in Feature Extraction Pro.

Workflow

The process of detecting poles can be split into four main steps:

  1. Raw input data interpretation
  2. Set unique parameters for pole detection
  3. Detect representative poles on small zones
  4. Detect poles for all input data

Interpret Data, Inspect Poles

Each data set is unique and as such a new data interpretation is required. Use the point cloud “Legend Editor” (RGB / Intensity / Distance / Height / Classification, etc.) to clarify pole visual detection in Map 2D or Map 3D. Pole detection is based on the geometry of the point cloud: poles are expected to be round and standing straight. Get the average diameter and height of the poles by using the measure tools (use “Free line” and “Point drop to ground”). You can also use the “Slice view” to help ascertain the diameter. The larger the diameter or/and the height, the greater the minimum number of points that define a feature. These values will be used in the pole detector parameter settings.

Set Parameters

Set the parameters in the “Pole Detector” settings according to the above-detected values (diameter, height, and number of points inside a detected feature). Optional: exclude trees, traffic signs, and poles which are not on the ground.

Run detection on Sample Area

In pole detector under “Area delimitation”, create a small rectangular region around a selection of poles. In the Pole Detector under “Analyze”, select “Start now”. This will conclude with two types of results: Detected and Candidates features. If satisfied with the outcome, you can save the results. The results will not be retained unless saved.

If the detection of all poles has not been successful, you can use the “Measure” tool under “Test” to click on any undetected pole (within our small region of interest) to adjust the overall parameters. Once completed, you can enter the new values under “Parameters” and start a new detection until all required objects are detected. Repeat the process until the required number of poles are detected and the outcome will be displayed under “Results”.

Run detection and handle results

Once you are satisfied with the results, under “Area delimitation” use the “Clear” button to delete the small rectangular region.

You will now have two choices for further processing:

  • Define an area of interest on your point cloud (our recommended option) as to eliminate confusion and to make the zone you are researching smaller (as to quickly exclude areas that cannot contain poles like vegetation or buildings, etc.) or
  • Run the process on the complete point cloud

The results can be imported and used to create an Inventory theme to further analyse the data using Manage Asset Inventory Themes and Verify inventory items.

 
Last modified:: 2022/04/15 06:19