If you do not see these features in your Neara projects and would like to explore Automatic LiDAR Classification, contact our team
Once you have imported LiDAR data into a project, Neara Pipelines make it easy to automatically classify the point cloud.
About Pipelines
Pipelines are a powerful feature of Neara. They allow you to create and run workflows that automate complex, large-scale network model data processing tasks.
The Auto-classifer pipeline is a pre-built pipeline optimized for the automatic classification of LiDAR data.
Open the Pipelines tab
Click the (+) button at the end of any group of tabs in the project workspace, and select Pipelines under the Tools section in the popup:
Any existing pipelines will be listed on the left-hand side of the tab. You can adjust the width of the divider to accommodate long pipeline names:
Add a new auto-classifier pipeline
Click the + New auto-classifier pipeline option.
A new auto-classifier type pipeline is added to your list of pipelines, with a default name. You can rename it by double-clicking it:
You can create many auto-classifier pipelines, each with their own configuration for different classification tasks
An auto-classifier pipeline has a standard configuration and stages. They are displayed on the right-hand side of the tab:
Configure the pipeline
The auto-classifier pipeline consists of five stages:
Input
Pre-processing
Classification
Post-processing
Output
Input
Click the Point cloud data source dropdown and select the point cloud dataset that you previously created from imported LiDAR data files:
Pre- and post- processing options
The Preprocessing and Postprocessing stages provide useful options and presets to normalize, denoise, and apply other filters to the data before and after the classification process. Use them as-is, or fine-tune them to suit your LiDAR source data.
One of the most common tasks in LiDAR classification is denoising. The Auto-classifier pipeline includes useful denoising presets:
For a deep-dive into how the denoise presets work, as well as tactics to help you tune them to suit your source LiDAR data, see How auto-classifier pipeline denoising presets work
For more control, expand the Advanced options inside one of these stages to explore and configure additional processing steps:
Output
In the Output stage, set the name of the output dataset that will be produced post-classification:
Preview the classification output
Neara offers a fast and easy way to preview the classification and effects of different pre- and post- processing options without needing to process your entire point cloud.
First, make sure that your pipeline has a Point cloud data source input, and an output Dataset name.
Select a limited area to preview by clicking the Select area option at the top of the Pipelines panel and then position the preview area selector outline over an area that contains LiDAR data, in the Perspective View:
The preview area will disappear if the Pipelines panel is not visible, or another pipeline is selected
If you select an area that does not contain any LiDAR data, the Preview option will be disabled
Once selected, click the Preview button. When the preview is ready the output will be updated in the Perspective View to show classified point cloud within the preview area only:
You will also notice that a visibility (eye)
icon is shown for each stage of the pipeline, and for each step within those stages.
You can toggle the visualisation of each individual stage and/or step to see how they affect the classification process before you start processing the entire dataset.
Classify the point cloud
If you are satisfied with the classification preview, you're ready to process the pipeline. You can classify the full LiDAR dataset, or just the selected area.
When you are ready, click Process. You will be presented with a confirmation dialog before processing commences:
To cancel pipeline processing, click the Stop button at the top of the panel:
During import, the progress bar on the Layers tab will display different colors:
Preparing resources: an animated / hashed grey bar that indicates Neara is preparing computing resources. Import and dataset processing has not yet started.
Processing: a blue bar that indicates import and dataset processing is underway. As processing continues, you will notice portions of the blue bar turn yellow and green. The yellow section indicates the portion of processing currently active, and the green section indicates the portion of processing that has completed.
Complete: a solid green bar indicates the import process is complete.
Queued: a solid grey bar indicates future stages queued for process, but that are not yet ready to start.
Use the classified point cloud
Once processed, your classified LiDAR dataset is shown in the Point Clouds section of the Layers panel.
To open the Layers panel, click the (+) button at the right hand end of the tabs in any section of the workspace and select Layers under the Tools section in the popup.
Click the visibility (eye)
icon next to the dataset name in the Layers panel to toggle the visualisation of the dataset in the Perspective View:
