Overview

Annotate images offline with Labelme

Unofficial preview. Docsbook assembled these pages from the public wkentaro/labelme README.md and examples/ directory. It is not affiliated with, endorsed by, or maintained by the Labelme project. The official documentation lives at labelme.io/docs.

You draw polygons, rectangles, circles, lines and points over an image, and Labelme saves them as a JSON file next to it. It is written in Python and uses Qt for its interface.

This preview exists to show one thing: what happens when the material that is already public — a single 12 KB README and six example folders — is split so that each task the reader arrives with has a page of its own, a title that names that task, and a next step.

Labelme showing an image with several labelled instance-segmentation polygons drawn over it

What this preview changes#

The underlying material is unchanged — every command, flag and file name below came from the repository. What changed is where it lives:

In the repository In this preview
One README with an ### FAQ bullet list 19 pages, each titled with the task it answers
Export instructions inside example folders Export to COCO and Export to Pascal VOC as their own pages
CLI flags described in a prose paragraph A table with every flag from labelme --help
No search, no next/previous Search, breadcrumbs, and a next step at the bottom of every page

Next steps#

Start with Get started if you want Labelme running, or Core concepts if you want to understand the annotation model first.

The real thing
Labelme's own site and downloads

This preview is a demonstration of layout. For the product itself, the current prices and the official docs, go to the source.

Updated

Was this page helpful?