Transloadit

What is Automatic Cropping?

Automatic cropping uses predefined rules or content analysis to select a crop region for a target aspect ratio. It aims to preserve important subjects while removing less relevant image areas.

How Automatic Cropping works

Automatic cropping chooses a source rectangle that matches a requested output shape, using geometry, detected subjects, saliency, faces, or stored focal points as guidance. It differs from scaling because pixels outside the selected rectangle are discarded rather than squeezed into new proportions. The chosen region may vary by target rendition. Image pipelines apply the crop before resizing, then preserve coordinates or previews for review and repeatability.

Key facts

  1. A crop computed for a wide banner should not be reused blindly for a square thumbnail, because the valid rectangle and subject placement change with the target ratio.
  2. Face or saliency detection returns evidence rather than editorial intent; group photos, text overlays, and off-center products are common compositions needing manual focal data.
  3. Applying orientation metadata after crop analysis can move the selected rectangle to the wrong area, so pipelines should normalize orientation before calculating coordinates.

When Automatic Cropping matters

Use automatic cropping to create responsive thumbnails when manual positioning would not scale. Subject detection can fail on unusual compositions, so important outputs may need focal-point overrides.

Turn media knowledge into a working pipeline

Connect uploads, processing, AI, storage, and delivery through one declarative API — with the encoding stack, scaling, and format churn handled for you.

Try Transloadit for free