LabelImg is a graphic image annotation tool. It is very easy to use and intuitive to operate, and it is a good helper for users who want to mark many labels! I wrote it in Python and uses Qt for its graphical interface. Labels are saved as XML files in PASCAL VOC format, which is the format used by ImageNet. Additionally, it supports YOLO and CreateML formats. We can store it in different formats according to user needs, let us see how to install and use it together!
Download windows_v1.8.1.zip, unzip it, and click labelImg.exe in the folder.
Go to the GitHub clone source file, or download the source compressed file and unzip it.
Open Terminal, first confirm whether Python3 and pip3 are installed, move to the labelImg directory and execute the following commands, if there is pip3, it is recommended to install pyqt5 and lxml with pip3.
Then execute make qt5py3, and open labelImg.py with python 3, as shown in the figure below.
Please refer to the Github.
Watch a demo video
Select the image storage directory.
Select the xml storage directory
If you have already marked the data, you will see the marked box after opening it. If the first one does not appear, press one and switch back to the previous one and it will appear.
Select the save format
According to training needs. (PascalVOC or YOLO)
|Ctrl + u||Load all of the images from a directory|
|Ctrl + r||Change the default annotation target dir|
|Ctrl + s||Save 儲存|
|Ctrl + d||Copy the current label and rect box|
|Ctrl + Shift + d||Delete the current image|
|Space||Flag the current image as verified|
|w||Create a rect box|
|d||Next image 下一張圖片|
|a||Previous image 上一張圖片|
|del||Delete the selected rect box|
|Ctrl++||Zoom in 放大|
|Ctrl–||Zoom out 縮小|
|↑ → ↓ ←||Keyboard arrows to move selected rect box|