Labelbox

Image Segmentation

A comprehensive set of image segmentation tools to efficiently segment partial or full images for computer vision applications

Overview

Watch this video tutorial to learn everything about the image segmentation tools in Labelbox.

Configuring Image Segmentation Interface

The interface configurator is on the left and the labeling environment is previewed on the right. Notice that you can add custom attributes per class

The interface configurator is on the left and the labeling environment is previewed on the right. Notice that you can add custom attributes per class

During project setup, select the segmentation interface:

From there you can setup your ontology by adding all of the objects and classifications needed for your project.

Pen Tool

The new Pen drawing tool is designed to be the fastest way to outline objects. It allows you to draw freehand as well as straight lines, similar to a polygon tool. You can also use the Pen tool to erase, just click the (-) icon in the top bar. Tip: Hold Alt on your keyboard to temporarily switch to the erase mode while you draw.

Layer Locking (Draw to back)

Easily lock existing labeled instances to draw regions behind them.

Easily lock existing labeled instances to draw regions behind them.

With layer locking, you are able to continue annotating new objects without overwriting the objects you have already drawn. In the example above, the fish have already been annotated. When the Water class is chosen and the draw to back option is selected the sections of water are annotated behind the fish which essentially locks them from being modified. This can significantly speed up labeling since it is not required to intricately outline around the border of other objects.

Steps:

  1. Select the object.
  2. Select the draw to back option.
  3. Begin drawing without fear of overwriting objects you have already annotated.

Creating Instances

Creating 2 instances of "Orange Fish"

Creating 2 instances of "Orange Fish"

You can create multiple instances of the same object for instance segmentation tasks. In the example above, you can see the first "Orange Fish" being drawn. Then a separate instance is initiated by selecting "Orange Fish" again, then continuing to draw the next fish.

Steps:

  1. Select a class and draw the object.
  2. Select the same class again and draw the next instance of the object.

Nested Classifications

Custom attributes per instance

Custom attributes per instance

If you have configured the interface with nested classifications for any of your objects then once you have completed annotating an object you will be presented with the classification questions. If an object has a nested classification then you will see a dropdown arrow to the left of the object name as seen in the screenshot above. Note: If the classification is set to be required then the label cannot be submitted until the required questions are answered.

Keyboard Hotkeys

Collapsible keyboard hotkey menu available on the top right

Collapsible keyboard hotkey menu available on the top right

We have designed our hotkeys to maximize the efficiency for labelers. There are hotkeys designated for selecting objects, zoom and viewing controls, drawing tools, and other core actions.

Label Format

"Label": {
    "objects": [{
      "featureId": "cjxtm2d32i9aa07940tifrpuh",
      "schemaId": "cjxtjkpjai8t80846iwqaa1d8",
      "title": "Orange Fish",
      "value": "orange_fish",
      "color": "#3F51B5",
      "instanceURI": "https://api.labelbox.com/masks/cjxtjbwiah4990721z4895xnt/cjhfn60b30snn0756xunst31j/cjxmdpg2zmbzw0cwb35b7br93/1?feature=cjxtm2d32i9aa07940tifrpuh&token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiJjamhmbjYwYjMwc25uMDc1Nnh1bnN0MzFqIiwib3JnYW5pemF0aW9uSWQiOiJjamhmbjV5NnMwcGs1MDcwMjRuejFvY3lzIiwiaWF0IjoxNTYyNTQ0MjIzLCJleHAiOjE1NjUxMzYyMjN9.b64mmRqUY8zVkTBihqSQXelbnKFyi-S_yAIYk7xVi8c",
      "classifications": [{
        "featureId": "cjxtm2e8zheyt0863sds45xyx",
        "schemaId": "cjxtjkphehbnj0848wknmg44u",
        "title": "Is the fish blurry?",
        "value": "is_the_fish_blurry?",
        "answer": {
          "featureId": "cjxtm2ebgi7tx0944gr8jx0lp",
          "schemaId": "cjxtjkpfwh4y90721uodxps2b",
          "title": "Yes",
          "value": "yes"
        }
      }]
    }, {
      "featureId": "cjxtm2kf0i9b207944z3gypsn",
      "schemaId": "cjxtjkpjai8t80846iwqaa1d8",
      "title": "Orange Fish",
      "value": "orange_fish",
      "color": "#3F51B5",
      "instanceURI": "https://api.labelbox.com/masks/cjxtjbwiah4990721z4895xnt/cjhfn60b30snn0756xunst31j/cjxmdpg2zmbzw0cwb35b7br93/1?feature=cjxtm2kf0i9b207944z3gypsn&token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiJjamhmbjYwYjMwc25uMDc1Nnh1bnN0MzFqIiwib3JnYW5pemF0aW9uSWQiOiJjamhmbjV5NnMwcGs1MDcwMjRuejFvY3lzIiwiaWF0IjoxNTYyNTQ0MjIzLCJleHAiOjE1NjUxMzYyMjN9.b64mmRqUY8zVkTBihqSQXelbnKFyi-S_yAIYk7xVi8c"
    }, {
      "featureId": "cjxtm2spbh8xx0721s18nwqq0",
      "schemaId": "cjxtjkpjai8t90846yzpbpo7t",
      "title": "Blue Fish",
      "value": "blue_fish",
      "color": "#F4511E",
      "instanceURI": "https://api.labelbox.com/masks/cjxtjbwiah4990721z4895xnt/cjhfn60b30snn0756xunst31j/cjxmdpg2zmbzw0cwb35b7br93/1?feature=cjxtm2spbh8xx0721s18nwqq0&token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiJjamhmbjYwYjMwc25uMDc1Nnh1bnN0MzFqIiwib3JnYW5pemF0aW9uSWQiOiJjamhmbjV5NnMwcGs1MDcwMjRuejFvY3lzIiwiaWF0IjoxNTYyNTQ0MjIzLCJleHAiOjE1NjUxMzYyMjN9.b64mmRqUY8zVkTBihqSQXelbnKFyi-S_yAIYk7xVi8c"
    }, {
      "featureId": "cjxtm3i3ci9cx07940e5t7if8",
      "schemaId": "cjxtk7z5qi5s90794v7wvseu6",
      "title": "Water",
      "value": "water",
      "color": "#EF6C00",
      "instanceURI": "https://api.labelbox.com/masks/cjxtjbwiah4990721z4895xnt/cjhfn60b30snn0756xunst31j/cjxmdpg2zmbzw0cwb35b7br93/1?feature=cjxtm3i3ci9cx07940e5t7if8&token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiJjamhmbjYwYjMwc25uMDc1Nnh1bnN0MzFqIiwib3JnYW5pemF0aW9uSWQiOiJjamhmbjV5NnMwcGs1MDcwMjRuejFvY3lzIiwiaWF0IjoxNTYyNTQ0MjIzLCJleHAiOjE1NjUxMzYyMjN9.b64mmRqUY8zVkTBihqSQXelbnKFyi-S_yAIYk7xVi8c"
    }],
    "classifications": []
  }

Image Segmentation


A comprehensive set of image segmentation tools to efficiently segment partial or full images for computer vision applications

Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.