Exploring data

Annotating images

Zegami's annotation tools let you mark up images directly in the viewer — region by region, with classes — and round-trip the result through the standard W3C Web Annotation format.

Last updated 2026-05-29

Opening the annotation tools

Click an image to open the Inspector, then switch to the Annotations tab. You get the drawing tools, a class picker, image adjustments, and the list of annotations on the current image.

The Inspector Annotations tab with image-adjustment sliders, Select/Rectangle/Polygon/Point tools, and a class picker
The Annotations tab: adjustment sliders, the drawing tools, and the active class.

Drawing tools

  • Select — the default; click an existing annotation to select, move, or delete it.
  • Rectangle — drag a bounding box.
  • Polygon — click to place vertices, close the shape to finish.
  • Point — drop a single marker.

Each new annotation is stamped with the active class (below the tools). Annotations are stored with normalized [0, 1] coordinates, so they stay correct regardless of image size or zoom.

Classes

The class picker sets what each new annotation means (for example “defect”, “cell”, “vehicle”). Manage the class list — add, rename, recolour — from the class-management control. Classes are collection-wide, so everyone annotating uses the same vocabulary and colours.

Image adjustments

The same tab carries display adjustments — brightness, contrast, saturation, hue, invert — so you can dial in detail before marking a region. For radiology collections this is where window/level lives. Adjustments are view-only; they don’t alter the stored image.

Tags

Tags are a lighter-weight label than annotations — a flat keyword on the whole image rather than a region. Add and remove them on the Inspector’s Information tab. Then filter by tag from the Filters panel — see Filtering & search.

Import and export (W3C)

Annotations round-trip through the W3C Web Annotation format:

  • Export them (alongside other formats) from the Export panel — see Exporting.
  • Import existing W3C annotations to overlay them on the matching images.

Because coordinates are normalized, annotations created elsewhere line up as long as they target the same images.

See also