Issues with Matterport Property Reports — Understanding the Causes and Improving Results
The introduction of the automated Property Report feature from Matterport has generated considerable interest across the property, architecture, engineering and facilities management sectors.
The report offers a fast and highly automated method of generating building outline plans and room data directly from Matterport digital twins. However, recent support discussions with purchasers of Matterport Pro3 cameras through Hitechniques Limited, combined with feedback from participants attending the CaptureTrain Certified Matterport Capture and Postproduction Training Course, have highlighted several recurring issues affecting report accuracy.
These include:
- Missing rooms
- Missing doors
- Incorrect room recognition
- Unexpected shapes appearing within plans
- Spaces being incorrectly categorised
Understanding the Matterport Property Report
The Matterport Property Report is a relatively new feature and remains an evolving technology.
The report is generated automatically using AI-assisted analysis of the spatial data captured during scanning. It is designed to provide:
- Outline floor plans
- Room dimensions
- Ceiling heights
- Approximate room areas
While the automation is impressive, the quality of the final output is directly linked to the quality of the original capture process.
In practice, the Property Report should be viewed as an intelligent automated interpretation of the captured data rather than a manually verified architectural drawing.
_1000.png)
Common Causes of Errors
Analysis of problematic models supplied by clients suggests that many issues originate during scanning rather than during report generation itself.
1. Missing Doors
One recurring issue involved doors disappearing from the generated outline plan.
In several cases, the operator had captured only a single scan position within a doorway rather than taking scans on both sides of the door threshold. This reduced the AI’s ability to interpret the doorway correctly within the spatial model.
2. “Spray” Data from Unmarked Windows
Another common issue involved windows that had not been properly marked during capture.
This allowed LiDAR or visual data to project outside the building envelope, creating what operators often describe as “spray”.
Consequences included:
- Rooms not being recognised correctly
- External artefacts appearing within plans
- Distorted room geometry
In one example, the external spray data caused the AI to generate a non-existent bay window within the Property Report.
3. Roller Shutters and Open Industrial Spaces
In a commercial building consisting of offices and warehouse areas, the warehouse had been scanned with full-height roller shutters open at both ends. The result was that the AI interpreted the space as an open-sided carport rather than an enclosed internal room.
This highlights the importance of controlling environmental conditions during capture wherever possible.
Labels and Schematic Floor Plans
Another point causing confusion relates to room labels.
Custom room labels added during Matterport Workshop postproduction can appear within Schematic Floor Plans when configured appropriately. However, these labels currently do not transfer into the automated Property Report.
This distinction is important for clients expecting consistency between the two products.
The Importance of Good Capture Practice
The key lesson emerging from these support cases is straightforward: AI-assisted outputs are only as good as the data supplied to them.
As Matterport continues to refine automated reporting tools, operators must increasingly focus on disciplined scanning methodology and understanding how environmental conditions affect AI interpretation. Professional capture practice remains essential.
This is one of the core themes explored during the CaptureTrain Certified Matterport Capture and Postproduction Training Course, where scanning methodology, troubleshooting, workflow optimisation and postproduction techniques are covered in detail