All language subtitles for 003 Dataset Annotation using LabelImg

Okay, Then we'll show you how to set up and use label image on Windows to begin open and control navigate

to by pressing the Windows Key and then tapping Anaconda.

Click on Anaconda Navigator.

If it's already open click environments.

To create environments.

Click, create, then enter the environments name.

In this example, we name the label IMG.

In Python, select Python 3.9 because label image could not be used in Python 3.10 when this course

was created.

Then click the button.

Wait until the environment creation is finished.

It will look like this when it is finished.

Then once the end on the prompt, press the windows key.

Then type.

Anaconda.

Click on any kind of prompt.

After that activate the environment, use the command activate label IMG.

If it looks like this, it means that the environment is active.

Use the following command to install the label image strip install label IMG.

Wait until the installation is finished.

When finished, it will looks like this.

Prepared the data set to be annotated next.

In this example, we will use the previously downloaded face mask dataset.

We never get to the downloads folder.

This is a face mask dataset.

There are two folders in this dataset images folder which stores images.

And the annotations folder which stores the annotated files because we want to show you how to do annotation.

Only the images will be used.

The next step is to extract the dataset in this video extract using tools from Windows 11.

To extract files using Windows 11 tools.

Right.

Click on the file and select.

Extract or.

Select the directory of order to save the extracted results.

In this case, you'll save to the directory.

Click extra.

Wait until the extraction is finished.

When finished, we go to the D directory.

The two extracted folders are shown below.

We'll delete this annotation because we're not using it.

The next step is to create a text file that contains the class names.

There are three classes in this face mask dataset.

Mask, no mask and that mask to create a new file.

Right click.

Then select new text document.

Name the file classes.

Open the file.

In a class one class.

Masks for those who use masks.

Class No mass for those who do not use a mask.

In red masks for those who use mask incorrectly.

Saved by pressing control as.

Move this file to the folder where the dataset is stored.

In this case, the image is folder.

Click the file, then press control X.

Then faced by pressing control fee.

We turn to any kind of prompt after that?

And then we get to the directory that stores the dataset.

Next, open the label image using the command.

Liberal emcee.

Then the folder that stores the dataset.

In this example, the images folder.

Then the file containing the class names.

In this example file classes.

Dot text in the images folder.

Then press the enter key.

The following is an initial view of the label image.

Next, we will annotate the image.

Click create checkbox to add an annotation.

Left click and read the mouse to create a bonding box that includes the object.

Then to a class.

After that, click.

Okay.

In addition, we can also add annotations by pressing the W key.

Press the key on the keyboard.

Left click and drag the mask to create a bonding box that includes the object.

Select the class, then click okay.

When you have finished annotating an image, click Save to save the annotation.

The YOLO annotation file has a text extension.

If the extension is still XML, click cancel.

First change the other format by pressing the following button.

After that we turn it to YOLO format by pressing the following button.

Click Save.

The name of this file is based on the name of the image.

Click Save to move to the next image.

Press the D key.

We annotate this image.

This is an example of an object with incorrect Massachusetts.

If you get it wrong or inaccurate when annotating, you can delete it.

For example, we will delete this box.

To delete it, click the box, then click the red box.

We can also delete annotations with the DEL Key.

Click the box, then press the Del Key.

Annotate all images in the dataset.

If you're finished, close, the label emits.

We have provided an annotated face mask dataset.

The dataset can be downloaded and used.

The data set is available for download at the following link.

See you in the next video.