How to Semi-Automatically Generate a Boundary Polygon

Overview

LP360 provides a semi-automated workflow for generating boundary (footprint) polygons that represent the true spatial extent of a point cloud. This approach offers a faster alternative to fully manual digitization while still allowing users to review and refine the final polygon as needed.

This workflow is well-suited for creating boundaries that closely follow the shape of the data, rather than simple rectangular extents.

Common Use Cases

This method can be used to generate boundary polygons for:

• Overall dataset extents (project footprints)
• Vegetation areas using classified vegetation points
• Water bodies, bridge decks, or building footprints after classification using LP360 AI Classification tools
• Any scenario where a polygon should represent the spatial coverage of a specific class or group of classes

When used with AI-derived classifications, this workflow allows you to quickly polygonize features such as water surfaces, structures, or buildings and then optionally refine them using standard feature editing tools.

Prerequisites

• LP360 with access to the required Point Cloud Tasks
• A LAS or LAZ dataset loaded into LP360
• LAS Layer must be open for read/write. 
• Data classified using traditional or AI Classification tools
• Ability to export and re-import LAS or LAZ files

Step-by-Step Workflow

Step 1: Thin the Point Cloud

Thinning reduces point density while preserving the overall shape of the data, making boundary extraction more reliable.

1. Open Classify by Statistics from the Point Cloud Tasks ribbon.
2. Apply thinning to a density appropriate for the level of detail required.
3. Optionally filter the input by class (for example, vegetation, water, bridge deck, or building classes produced by AI classification).

Step 2: Export the Thinned Dataset

Export the thinned point cloud to create a simplified dataset for boundary generation.

1. Open the Export Wizard.
2. Export the thinned points to a new raster, LAS or LAZ file.
3. Load the exported file back into LP360.

Step 3: Run Point Cloud Tracing and Squaring

Use Point Cloud Tracing and Squaring to automatically generate polygons from the thinned dataset.

1. Open Point Cloud Tracing and Squaring from the Point Cloud Tasks ribbon.
2. Use the dropper option to help determine an appropriate Grow and Trace Window or set the following parameters as a starting point:
   • Maximum Grow Window Area: Large enough to encompass the full dataset extent
   • Grow and Trace Window: Generally, if you set this value by sampling the data with the eyedropper tool, the resulting value will be correct. This is achieved by pressing the eye dropper tool and drawing a sample polygon in the area to be traced. The polygon is drawn by left clicking at each desired vertex and double clicking to end.
3. Run the task to generate one or more boundary polygons.

Step 4: Review and Refine the Polygon

After polygon generation:

• Review the resulting boundary feature or features.
• If multiple polygons are created:
  • Increase the grow and/or trace window parameters and rerun the task, or
  • Manually delete, merge, or reshape polygons using Feature Edit tools.

This workflow is intentionally semi-automated, allowing you to balance efficiency with control over the final geometry.

Parameter Tuning Quick Reference

Related AI Classification Workflows

AI Water Classification

LP360 AI Classification tools can identify and classify water surfaces in point cloud data. Once water points are classified, they can be filtered, thinned, and traced to generate water body polygons such as rivers, lakes, ponds, and reservoirs.

AI Building Classification

LP360 AI Classification can automatically identify building-related features, including building roofs and associated structural elements. These classified building points can be used to generate building footprint polygons suitable for mapping, planning, or GIS export.

Notes and Limitations

• This method does not create swath-level or flight-line footprints.
• Results depend on thinning density, class selection, and grow and trace window parameters.
• Manual QA/QC and editing are recommended before using the polygon as a final deliverable.
• Known issue: Procedure only works for classes other than 2 and 8.

Summary

By thinning a point cloud and applying Point Cloud Tracing and Squaring, LP360 users can efficiently generate boundary polygons that closely follow the true shape of their data. When combined with AI Classification tools, this workflow supports rapid polygon creation for vegetation, water, bridge decks, buildings, and other classified features while still allowing full manual refinement when needed.