/
Edge Detection

Edge Detection

Menu: Analyse / Edge detection

The Edge Detection tool automatically identifies and vectorizes edges in your point cloud, such as those found in buildings, train stations, and other structures. The tool generates points along detected edges, provided that the set parameters are met.

Note: This function requires a high density of points to ensure accurate edge detection.

1. Edge Detection Parameters

Search Radius – Defines the maximum radius within which the algorithm searches for the next point.
Tolerance Deviation (°) – Determines the maximum deviation between points:

  • Lower values → Keeps edges straighter.

  • Higher values → Allows for curved edge detection.
    Accuracy – Defines the maximum elevation difference from the edge for points to be considered.

  • Must be kept low for the algorithm to function correctly.

2. Detection Modes

The tool offers three automatic detection modes:

By One Point – Click a single point in the point cloud, and detection occurs on both sides of the selected point (similar to catenary detection).
By Two Points – The first point marks the start of detection, while the second point defines the direction the extraction will follow.
By Multiple Points – Similar to By One Point, but instead of detecting one edge at a time, the software will detect all edges at once (one edge per point clicked).

3. Running the Detection

1️⃣ Select a Detection Mode.
2️⃣ Click “Pick Points” and select an edge in the point cloud to provide a starting point.
3️⃣ Once the detection is complete, choose from the following options:

Save – Saves the detected edges to the VisionLidar project.
Load – Loads a previously saved edge extraction back into the project.
Append – Adds new edge detections to an existing extraction.
Export – Exports extracted edges in various formats:

  • .DXF (CAD)

  • .SHP (Shapefile)

  • .MID

  • .GeoJSON

  • .KML (Google Earth)
    Create Chains – Opens the “Enter Points/Chains” window to apply advanced processing functions (see Enter Points/Chainssection).

                                                                                                      

 

Related content