Automatic people masking is an Aura 2.2 feature that detects and removes people from colorized point clouds and 360 panoramic images during processing. It is enabled by default and requires no manual template selection. To adjust the settings, see Configure automatic people masking.
What is it?
Automatic people masking is an Aura processing feature that detects people in the source images captured by an Emesent scanner and removes them from the colorization and 360 panoramic image outputs. The feature replaces or augments the existing operator masking templates, which use fixed regions to mask a known operator position.
Detection runs against the mission's source images during processing. Aura identifies people regardless of whether the person is the operator carrying the scanner, a colleague in frame, or a bystander present in the captured environment. No manual selection or template adjustment is required.
Detection coverage
Automatic people masking detects people only. Vehicles, equipment, signage, and other objects are not masked. Use a manual mask template, or post-process the output, to remove non-person content.
Operator masking templates
Operator masking templates (right-handed, left-handed, centre) remain available alongside automatic people masking. Use a template when automatic detection is disabled, or when the operator position is fixed and known.
Why it matters
Automatic people masking reduces the manual effort required to produce clean, shareable mission outputs and supports privacy and compliance requirements when publishing deliverables.
Key benefits
Removes operator and bystander faces from colorized point clouds without manual selection.
Blurs people in 360 panoramic images so deliverables can be shared without revealing identities.
Reduces the need to plan operator position and scanner-handedness around a fixed mask template.
Detects multiple people in a single frame, not just the operator.
How it works
During Colorization or Extract 360 images processing, Aura analyzes each source image and identifies regions containing people. The identified regions are then excluded from the colorization output and masked or blurred in the 360 panoramic image output, depending on the configured settings.
The feature is enabled by default in Aura 2.2 for both the Colorization and Extract 360 images workflows. Detection runs on every source image associated with the mission, including images captured at operator head height.
Expected outcomes
The output produced by automatic people masking differs between the two workflows.
Colorization output
The colorized point cloud excludes points that fall within regions identified as people. The result is a point cloud free of operator and bystander imagery, with no face smearing across surfaces near the scanner.
Points that fall outside detected regions are colorized normally using the source image data.
Extract 360 images output
Each output 360 panoramic image is masked across regions identified as people. By default, a solid mask is applied. When Blur masked regions is enabled, a blur is applied instead, preserving the surrounding image context while obscuring identifying features.
Images that contain no detected people are produced without any mask applied.
Common use cases
Automatic people masking applies to most field workflows where the operator or other people are visible during the mission. The two scenarios below describe the most frequent cases.
Scanning with operator visible to the scanner
The operator carrying an Emesent scanner is regularly visible in the source images, particularly in images captured at head height. Automatic people masking removes the operator from both the colorized point cloud and the output 360 images without requiring a handedness-specific template.
For the configuration procedure, see Configure automatic people masking.
Scanning environments with bystanders
Active worksites, public spaces, and shared facilities frequently contain people other than the operator. Automatic people masking detects and removes these people in addition to the operator, supporting workflows where deliverables are shared with external parties.
Important considerations
Detection accuracy depends on lighting, distance from the scanner, and how much of the person is visible in each source image. Partially occluded people may not be detected in every frame.
Processing time may increase when automatic people masking is enabled, depending on the number of source images and the size of the mission.
The feature requires Aura 2.2 or later. Earlier versions do not support automatic detection and rely on manual operator masking templates only.