AI Upscaling / Interpolation

Category: Technology

Methods that computationally enlarge an image, inventing additional pixels. The result looks more detailed but contains no genuinely captured detail.

Interpolation and AI upscaling computationally increase an image's pixel count: missing image points are estimated, and with AI methods plausibly added from learned patterns. The image appears larger and seemingly sharper.

Crucially, this detail is invented, not captured. On close inspection artefacts, false structures or a smoothed-over impression appear. True gigapixel images are created purely optically by stitching real frames – without AI upscaling. That is precisely what distinguishes them from upscaled images.

Find images matching this topic via visual search in the portal.

Browse all images →

Explained by Daniel Richter · Expert in gigapixel photography · Gigapixel GmbH