🤖Supervisely JSON

Overview

Easiest way to import your images with annotations is to use the Supervisely format. Check out the Supervisely JSON format documentation for more details.

The Supervisely json-based annotation format supports such figures: rectangle, line (polyline), polygon, point, bitmap (mask), graph (keypoints), alpha mask, 2D cuboid. It is a universal format for various task types and is used in the Supervisely platform.

Enterprise users have access to "Import as links" option, which supports import of this format with annotations. This option might be beneficial in many cases, as it allows data import to Supervisely platform without re-uploading, maintaining a single source and speeding up import process.

To step up import speed even further you can compress all annotation files (.json's) into an archive and import it together with the images. (Note: This method is format-dependent and may not apply to all formats.)

Format description

Supported image formats: .jpg, .jpeg, .mpo, .bmp, .png, .webp, .tiff, .tif, .jfif, .avif, .heic, and .heif With annotations: Yes Supported annotation file extension: .json. Grouped by: Any structure (will be uploaded as a single dataset)\

Input files structure

Example data: download ⬇️.

Both directory and archive are supported.

Recommended directory structure:

  📦input_folder
   ┣ 📂dataset_name_01
   ┃  ┣ 📂ann
   ┃  ┃  ┣ 📄IMG_0748.jpeg.json
   ┃  ┃  ┗ 📄IMG_8144.jpeg.json
   ┃  ┣ 📂img
   ┃  ┃  ┣ 🏞️IMG_0748.jpeg
   ┃  ┃  ┗ 🏞️IMG_8144.jpeg
   ┃  ┗ 📂meta (optional)
   ┃     ┣ 📄IMG_0748.jpeg.json
   ┃     ┗ 📄IMG_8144.jpeg.json
   ┗ 📄meta.json

Project meta file meta.json is recommended to be present in the project directory. It contains classes and tags definitions for the project. If it is not present, app will try to create it from the annotations (if possible). Learn more about the meta.json file here.

Single-Image Annotation JSON

For each image, we store the annotations in a separate json file named image_name.image_format.json with the following file structure:

{
  "description": "food",
  "name": "tomatoes-eggs-dish.jpg",
  "size": {
    "width": 2100,
    "height": 1500
  },
  "tags": [],
  "objects": []
}

Fields definitions:

  • name - string - image name

  • description - string - (optional) - This field is used to store the text we want to assign to the image. In the labeling intrface it corresponds to the 'data' filed.

  • size - stores image size. Mostly, it is used to get the image size without the actual image reading to speed up some data processing steps.

  • width - image width in pixels

  • height - image height in pixels

  • tags - list of strings that will be interpreted as image tags

  • objects - list of objects on the image

Image annotation example with objects and tags

Example:

{
  "description": "",
  "tags": [
    {
      "id": 86458971,
      "tagId": 28283797,
      "name": "like",
      "value": null,
      "labelerLogin": "alexxx",
      "createdAt": "2020-08-26T09:12:51.155Z",
      "updatedAt": "2020-08-26T09:12:51.155Z"
    },
    {
      "id": 86458968,
      "tagId": 28283798,
      "name": "situated",
      "value": "outside",
      "labelerLogin": "alexxx",
      "createdAt": "2020-08-26T09:07:26.408Z",
      "updatedAt": "2020-08-26T09:07:26.408Z"
    }
  ],
  "size": {
    "height": 952,
    "width": 1200
  },
  "objects": [
    {
      "id": 497521359,
      "classId": 1661571,
      "description": "",
      "geometryType": "bitmap",
      "labelerLogin": "alexxx",
      "createdAt": "2020-08-07T11:09:51.054Z",
      "updatedAt": "2020-08-07T11:09:51.054Z",
      "tags": [],
      "classTitle": "person",
      "bitmap": {
        "data": "eJwBgQd++IlQTkcNChoKAAAADUlIRF",
        "origin": [535, 66]
      }
    },
    {
      "id": 497521358,
      "classId": 1661574,
      "description": "",
      "geometryType": "rectangle",
      "labelerLogin": "alexxx",
      "createdAt": "2020-08-07T11:09:51.054Z",
      "updatedAt": "2020-08-07T11:09:51.054Z",
      "tags": [],
      "classTitle": "bike",
      "points": {
        "exterior": [
          [0, 236],
          [582, 872]
        ],
        "interior": []
      }
    }
  ]
}

Last updated

Was this helpful?