How to identify differences between 2 point cloud - Proximity Classifier PCT

Overview

This workflow explains how to detect differences between two point clouds captured over the same area at different times by using Proximity Classifier PCT. The goal is to compare an Older LAS file with a Newer LAS file and identify points in the newer dataset that do not match the older one within a defined tolerance.

This is useful for detecting:

• Ground changes
• New construction
• Removed structures
• Terrain modifications
• Other localized changes between two surveys

Prerequisites

Before starting, make sure:

• You have two point clouds of the same area
• Both point clouds are from different dates

For clarity in this workflow, name the datasets: Base LAS and ObjectLAS

Workflow

1. Load the two point clouds in LP360
Open the Base LAS (older) and Object LAS (newer) in the project. These should represent the same area, captured at different times.

2. Filter and classify the data to compare. Filter the datasets to isolate the classes you want to compare.
Examples: Ground, Buildings or Ground + Buildings
Make sure the selected classes are classified consistently in both LAS files.

3. Make the Newer LAS as Active Layer

4. Open Proximity Classifier PCT

Use the following settings.

Reclassify dataset Active Layer → Object LAS
Input Points Filter -->Select the classes to compare in the Object LAS

Proximity Points Select the Base LAS
Proximity Filter --> Select the classes to compare in the Older LAS

Dimension → 3D
Distance Method → Radius ; Value → 0.3 meters

Reclassify Action
Point Located → Reclassify To -->Choose an output class in the Object LAS, for example: Class 20
Point Not Located --> No Action

5. Perform the PCT

What did we do?

The tool checks whether each point in the Object LAS has a corresponding point in the Base LAS within a 0.3 m radius in 3D space.

• If a match is found, the point is reclassified to the chosen output class, such as Class 20
• If no match is found, no action is taken

This allows you to isolate points in the newer dataset that are spatially different from the older dataset.

6. Open Cluster by Distance PCT

Use the following settings:

Input LAS Layer → Object LAS
Input Points Filter --> Class 20

Output LAS Folder --> <LP360_PROJECT_PATH>\Cluster_LAS
Autodelte output LAS Files: Checked
Centroid File --> <LP360_PROJECT_PATH>\Cluster.SHP

Cluster separation → 5
2D
Min Point Count / Cluster: 100
Cluster Type Code: 1

7. Perform the PCT

What did we do?

The tool search for cluster of points with at least 100 points, and create a point. This points are created in a shapefile and can be used to identify differences between the 2 point cloud. The output is generated as a shapefile, but it can be exported as a KML too.
An alternative workflow could be to trace the cluster of points to create poligons showing the differences.