Objects binding

This guide explains how to bind (group) objects on images


For some labeling scenarios, it is needed to group objects. Let's consider the case when you need to label object parts and then group them together. Such binding will help you distinguish parts of different objects. In this tutorial, you will learn how to programmatically group objects together (binding) and how to work with existing bindings.
Everything you need to reproduce this tutorial is on GitHub: source code and demo data.
In this tutorial, we will create binding for the labeled parts of a single car:
Imput image and masks
This tutorial consists of two parts:
Part 1. Create labels with binding and upload them to Supervisely server.
Part 2. Methods needed to work with existing bindings.

How to debug this tutorial

Step 1. Prepare ~/supervisely.env file with credentials. Learn more here.
Step 2. Clone repository with source code and demo data and create Virtual Environment.
git clone
cd tutorial-object-binding
Step 3. Open repository directory in Visual Studio Code.
code -r .
Step 4. change ✅ workspace ID ✅ in local.env file by copying the ID from the context menu of the workspace. A new project with demo data will be created in the workspace you define:
CONTEXT_WORKSPACEID=619 # ⬅️ change value
Copy workspace ID from context menu
Step 5. Start debugging src/

Part 1. Create labels with binding and upload them to server

Import libraries

import os
from typing import List
from dotenv import load_dotenv
import cv2
import uuid
import supervisely as sly

Init API client

Init API for communicating with Supervisely Instance. First, we load environment variables with credentials and workspace ID:
api = sly.Api()

Create project

Create empty project with name "tutorial-bindings" with one dataset "dataset-01" in your workspace on server. If the project with the same name exists in your dataset, it will be automatically renamed (tutorial-bindings_001, tutorial-bindings_002, etc ...) to avoid name collisions.
workspace_id = int(os.environ["CONTEXT_WORKSPACEID"])
project = api.project.create(workspace_id, name="tutorial-bindings", change_name_if_conflict=True)
dataset = api.dataset.create(, name="dataset-01")
print(f"Open project in browser: {project.url}")

Define annotation classes

Let's define annotation classes and upload them to our new project on server:
class_car = sly.ObjClass(name="car", geometry_type=sly.Bitmap, color=[255, 0, 255])
meta = sly.ProjectMeta(obj_classes=[class_car])
api.project.update_meta(, meta)

Upload demo image to server

image_path = "images/image.jpg"
image_name = sly.fs.get_file_name_with_ext(image_path)
image_info = api.image.upload_path(, image_name, image_path)
print(f"Image has been successfully uploaded: id={}")
# will be needed later for creating annotation
height, width = cv2.imread(image_path).shape[0:2]

Read masks and create sly.Annotation

More details about how to create labels can be found in this tutorial.
# create Supervisely annotation from masks images
labels_masks: List[sly.Label] = []
for mask_path in [
# read only first channel of the black-and-white image
image_black_and_white = cv2.imread(mask_path)[:, :, 0]
# supports masks with values (0, 1) or (0, 255) or (False, True)
mask = sly.Bitmap(image_black_and_white)
label = sly.Label(geometry=mask, obj_class=class_car)
ann = sly.Annotation(img_size=[height, width], labels=labels_masks)

Create bindings

We know that all three masks are parts of a single car object. Let's bind them together. It is important to notice that any unique string can be label's binding_key.
key = uuid.uuid4().hex # key can be any unique string
for label in ann.labels:
label.binding_key = key

Upload annotation with binding to server

api.annotation.upload_ann(, ann)
As a result, we will have three objects of class car grouped together:
Result annotation with binding

Part 2: Work with existing binding

Download annotation

Let's download existing annotation (we created it in part 1) from server.
project_meta = sly.ProjectMeta.from_json(api.project.get_meta(
ann_json = api.annotation.download_json(
ann = sly.Annotation.from_json(ann_json, meta)

Access to binding keys

If binding_key is None then the label does not belong to any group.
print("Labels bindings:")
for label in ann.labels:
The output will be the following:
Labels bindings:

Access all binding groups in annotation

groups = ann.get_bindings()
for i, (binding_key, labels) in enumerate(groups.items()):
if binding_key is not None:
print(f"Group # {i} [key={binding_key}] has {len(labels)} labels")
# Binding key None defines all labels that do not belong to any binding group
print(f"{len(labels)} labels do not have binding")

Discard binding

Let's remove bindings on objects of class car:
for label in ann.labels:
if == "car":
label.binding_key = None
Or you can discard binding for all labels in annotation:
Let's upload updated annotation (without bindings) back to the server:
api.annotation.upload_ann(, ann)