Create a 3D model without 3D modeling

By the Visuality team · Published on September 19, 2023 · Updated on October 15, 2026 · 6 min read

In the example above, the armchair shown on the right was used as a reference to create the 3D model displayed on the left in augmented reality — all in just a few minutes and without any prior modeling!

Prerequisites

• The device used should have a high-quality camera sensor.
• The room should be evenly lit, without shadow areas or harsh contrast.
• An object you can reach from every side, ideally matte and textured.
• A Visuality account with a free object slot in your plan.

Download the RealityScan Mobile app:

Step 1: capture the object with RealityScan

Open the app and follow the instructions to create an account. You'll arrive at this screen:

Tap the "+" button to reach this screen:

Select "Augmented Reality Mode". You are now in camera mode.

You'll need to capture photos of the object from every angle, both close-up and from a distance. To improve the quality of the 3D model, follow these guidelines:

• Make sure photos are sharp and the object is photographed from all angles.
• Alternate between close-up shots for details and wider shots for the overall shape.
• Generally, 100 to 150 photos are needed depending on the object’s size.

Tap the red button to begin capturing, and move slowly around the object. The photo count appears at the bottom left. After some time, a point cloud will appear. Red areas indicate parts with insufficient coverage — focus on those, but don’t ignore the others.

Once you’ve taken enough photos, tap the arrow. You’ll reach the following screen:

An analysis is in progress, with a percentage shown on the right. Once it’s done, click the arrow at the bottom right.

On this screen, crop the object by adjusting the arrows. Use the handles to refine your selection.

Once the object is centered, click "Ready to process".

Choose the options shown in the screenshot, then click "Process Project".

Wait a few minutes for the object to be generated, then click on it.

You’ll see your object rendered. One last step: remove the floor by clicking the button on the far left.

As before, crop the object to remove its surroundings. Raise the rectangle slightly to remove the floor, then click "Crop".

Your object is now ready! Click "Export 3D model" to import it into Visuality.

Step 2: upload the 3D model to Visuality

Go to your 3D object library then click the "Add Object" button.

Select the option "I already have my 3D object".

You will then arrive on the page for adding a 3D model.

After validation, you’ll return to your library.

Step 3: view the object in augmented reality

Click "View/Share". You arrive at the object sharing page.

Two viewers are available:

Web Viewer: Accessible via a simple URL, no installation required.

App Viewer: Available via the Android app, ideal for large models and offline use.

You can generate a QR code for each viewer to easily share the model or integrate it into your website.

Good to know and limitations

• Please note that this method is not suitable for all types of objects. Some, especially highly detailed ones, may not render optimally and could require traditional 3D modeling.
• Shiny, transparent, very dark or plain surfaces are hard to reconstruct with photogrammetry.
• The uploaded ZIP file must contain the model in GLB format: Visuality extracts it and converts it with its textures.
• Using another scanning app, such as Polycam? Export the model as GLB or USDZ, then upload it in the 3D file field. Check which export formats are included in your plan for that app.
• After uploading, check the object's size: a scan is not always exactly at real-world scale.

Frequently asked questions

How many photos do you need to scan an object in 3D?

Usually 100 to 150 photos depending on the object's size, taken all around it, close up for details and from further away for the overall shape. The RealityScan app highlights poorly covered areas in red: fill them in before starting the model processing.

Which objects are hard to scan in 3D?

Shiny, transparent, very dark or plain-coloured objects are a problem for photogrammetry, which needs visible detail to rebuild the surface. For these products, custom modelling, which Visuality offers on quotation, or AI 3D generation from a photo may give better results.

Can I use an app other than RealityScan?

Yes. Any scanning app that exports to GLB or USDZ will do, for example Polycam, depending on the formats included in your plan. In that case, upload the file in the usual 3D file field rather than the RealityScan field, which expects a ZIP archive.

Is 3D scanning more accurate than AI generation?

Generally, yes: scanning rebuilds the object from what the camera actually saw, from every angle, whereas AI infers the hidden sides from one to four photos. On the other hand, scanning takes more capture time and the object must be physically available.

Questions? Feel free to contact us at contact@visuality.fr!

---

Related articles