“Development Technology” LabelImg installation and use introduction

LabelImg is a graphical image annotation tool.

It is written in Python and uses Qt as its graphical interface.

Annotations are saved as XML files in PASCAL VOC format, which is the format used by ImageNet. It also supports the YOLO format

Watch the demo video

The installation

Ubuntu Linux

Python 2 + Qt4

sudo apt-get install pyqt4-dev-tools
sudo pip install lxml
make qt4py2
python labelImg.py
python labelImg.py [IMAGE_PATH] [PRE-DEFINED CLASS FILE]
Copy the code

Python 3 + Qt5 (Recommended)

sudo apt-get install pyqt5-dev-tools
sudo pip3 install -r requirements/requirements-linux-python3.txt
make qt5py3
python3 labelImg.py
python3 labelImg.py [IMAGE_PATH] [PRE-DEFINED CLASS FILE]
Copy the code

brew install qt qt4
brew install libxml2
make qt4py2
python labelImg.py
python labelImg.py [IMAGE_PATH] [PRE-DEFINED CLASS FILE]
Copy the code

brew install qt  # Install qt-5.x.x by Homebrew
brew install libxml2

or using pip

pip3 install pyqt5 lxml # Install qt and lxml by pip

make qt5py3
python3 labelImg.py
python3 labelImg.py [IMAGE_PATH] [PRE-DEFINED CLASS FILE]
Copy the code

brew install python3 pip3 install pipenv pipenv --three # or pipenv install pyqt5 lxml pipenv run pip install pyqt5 lxml  pipenv run make qt5py3 python3 labelImg.py [Optional] rm -rf build dist; python setup.py py2app -A; mv "dist/labelImg.app" /ApplicationsCopy the code

Windows

Install Python, PyQt5 and install LXML.

Open CMD and go to the labelImg directory

Pyrcc4 -o line/resources.py resources.qrc for pyQt5, Py QRC python labelImg. Py python labelImg. Py [IMAGE_PATH] [pre-defined CLASS FILE]Copy the code

conda install pyqt = 5
pyrcc5 -o libs / resources.py resources.qrc
python labelImg.py
python labelImg.py [IMAGE_PATH] [PRE-DEFINED CLASS FILE]
Copy the code

Get from PyPI but only python3.0 or higher

Pip3 install labelImg labelImg labelImg [IMAGE_PATH]Copy the code

-e DISPLAY = Unix $DISPLAY \ --workdir = $(PWD) \ --volume = "/ home / $user: /home/$USER "\ --volume =" /etc/group: /etc/group: ro "\ --volume =" /etc/passwd: /etc/passwd: Ro "\ --volume =" /etc/shadow: /etc/shadow: Ro \ - volume = ""/etc/sudoers. D: / etc/sudoers. D: ro" \ - v/TMP/X11 - Unix: / TMP/X11 - Unix \ tzutalin/py2qt4 make qt4py2 ; ./labelImg.pyCopy the code

usage

Steps (YOLO)

1. indata/predefined_classes.txtDefines a list of classes that will be used for training.
2. Build and start using the above instructions.
3. Click the “PascalVOC” button at the bottom right of the toolbar under the “Save” button to switch to YOLO format.
4. You can use Open/OpenDIR to process single or multiple images. When the individual image is complete, click Save.

TXT files in YOLO format will be saved in the same folder as images with the same name. A file named “classes.txt” is also saved in this folder. “Classes.txt” defines the list of class names referenced by the YOLO tag.

Note:

• Your label list must not change during the processing of the image list. When an image is saved, classes.txt is also updated, while previous annotations are not.
• The “Default Class” feature should not be used when saved in YOLO format and will not be referenced.
• When saving to YOLO format, discard the “Difficult” flag.

hotkey

Verification picture:

When pressing the space bar, the user can mark the image as verified, which will show a green background. This is used when the dataset is automatically created, and the user can then browse through all the images and mark them instead of annotating them.

Difficult:

The difficulty field set to 1 indicates that the object has been annotated as “difficult,” for example, an object that is clearly visible but hard to recognize without taking full advantage of the context. Depending on your deep neural network implementation, you can include or exclude difficult objects during training.

license

Free software: MIT license