Data Processing Using Artificial Intelligence

At Nektar 3D Consulting we employ Artifical Intelligence to effciently process your point cloud data

Collecting more data than you can process?


Our consulting team utilizes artificial intelligence (AI) to help you extract high quality features from your LiDAR datasets in a highly efficient manner.

Independent of location, scanner, or collection method, our AI algorithms can help you efficiently classify and clean your point cloud data in a fully automated unattended manner.

 Extend the capabilities of your LiDAR team with accurate classified data.


Refine AI systems which accurately classify your LiDAR data at scale


Whether ground-based or aerial, rural or urban, our system adapts to your environment to deliver classified data with best-in-class accuracy

Regular Updates

We partner to gain access to datasets from around the world, getting up-to-date scans with high quality annotations to keep our system ready for changing environments

RGB-Colourized Point Cloud

Fully Classified Point Cloud

How it works

Our algorithms utilize principles of machine learning and AI to automatically classify every point within the imported point cloud into a specific type of object.

Our basemodels are trained to recognize multiple types of objects at varying levels of accuracy.

See the list below for examples of objects that the models have been trained to classify.

Roadway colored with RGB from cameras
Roadway colored with classification labels

Our Model’s Capabilities



Low/Medium/High Vegetation

Ground Elements

Road Surface



Lane Markings


Off Ground

Traffic Signs



Electrical Boxes

Construction Equipment

Concrete Barriers






Scanning Artifacts



Different Road Environments

Million Points


RGB-Colourized Point Cloud

Fully Classified Point Cloud

Vegetation Management

Utility Line Conflict Assessment

Data Compatability

Our models are trained to process data regardless of the scanning system or the sensor that was used to collect the data.

Scanners Scan Type
Trimble Mobile LiDAR
Leica Aerial LiDAR
Teledyne Optech

Although higher point density results in more accurate results. The algorithms can be customized to process data with low point density

Roadway from low density scanner - points only
Roadway from low density scanner - colored by classification label

Applications and Use Cases

Guardrail Assessment

Roadway showing classification label
Roadway hidden showing only classified guardrail

Noise and Vehicle Removal

Roadway with vehicle noise
Roadway with vehicle noise removed

Utility Powerline Extraction and Encroachment Analysis

Roadway showing classification label
Roadway hidden showing only poles

Extraction of Bridge Structure and Other Elements

Roadway with all
Roadway with bridge only

Extracting Multiple Features

Interested in a Free Demo?

Send us samples of your data today and we will send you the processed files for free.


Our team is also capable of annotating and training a model that is tailored to your datasets for improved accuracy.