Manually Label Object Detection Dataset With Bounding Boxes

This tutorial guides you through manually labeling an object detection dataset. You will learn how to create labeling tags, draw bounding boxes, and assign tags to objects for accurate dataset annotation.

1. Introduce Object Detection Labeling

Today, you will learn how to label an object detection data set. This is what we'll find the first time we open the labeling section in a dataset.

2. Create Labeling Tags

Click the Next button to continue setting up your labeling tags.

3. Enter Tag Label Name

Write a tag name you want to use in your dataset.

4. Add Additional Tag

Click the Add another tag button to include more tags for your dataset.

5. Create Labeling Tags

We just press Create to finalize the tag setup.

6. Tags Created

And we have the tags created.

7. Start Drawing Bounding Box

Now, with the tool called Draw bounding box selected, we can start drawing the bounding boxes.

8. Draw Bounding Box

Press one corner and then drag to the opposite corner to create a bounding box.

9. Select Tag for Bounding Box

In this case, the tag is a label not a barcode so we simply change it.

10. Zoom to Adjust Bounding Box

We can zoom in using Ctrl + Mouse Scroll or the "+" button in the top menu to get a closer view.

11. Correct Bounding Box

We use this tool to move the bounding box to the desired position.

12. Adjust Bounding Box Position

With this tool, press the desired bounding box to adjust its position.

13. Fixed

14. Continue for all detections and images

Keep labeling all the images until you’re finished!
Remember, this is manual labeling, don’t waste your time doing everything by hand. Use the available tools to make your work easier: Smart Labeling, Auto Labeling and Inference.