Photo Scan Common Failure Reasons and Tips for Improvement

If you used Photo Scan and obtained subpar results with surfaces scattered all over the place, it's most likely that either the photos were not up to standards, or that the objects possessed shiny/reflective/transparent surfaces.

The technology behind Photo Scan, photogrammetry, is an algorithm that stitches together unique feature points across images to reconstruct 3D meshes. These 'feature points' are crucial for the photogrammetry process, and they are usually the corners, edges, and bumps of objects that can be recognized by computer vision.

Typically, objects with rich feature points are rocks, wood, and sculptures. Conversely, objects that lack feature points can typically be water bottles, cars, etc.

Remember that Photo Scan is an effective scanning method only for those objects with lots of feature points, and it's best to avoid scanning shiny/reflective/transparent objects with Photos Scan.

To effectively use Photo Scan, consider the following tips:

• Choose the right object, it's recommended always to pick a featureful object rather than a featureless one.
• Take photos orbiting around the object.
• The phototaking orbits should be from the top, middle, and lower perspectives.
• Ensure 70% overlap between each photo taken.
• Try avoid moving the object throughout the capture.
• If you absolutely had to move the object during capturing, make sure to turn on Auto Object Masking.

With these tips in mind for your next scan, we're sure that you'll get a much better result.