PPOCRLabel is a powerful and efficient tool for labeling optical character recognition (OCR) data. Developed as part of Baidu’s PaddleOCR project, it is a user-friendly platform for annotating text in images, making it invaluable for training custom OCR models. This guide will walk you through installing PPOCRLabel on Windows, ensuring you can start your OCR projects quickly.

Prerequisite

Make sure your Windows runs on X-86 or AMD. If you have an ARM-based Windows, such as Robo & Kala, you won’t be able to install PaddleOCR and PPOCRLable on it.

Step1: Install Python3.9

Unfortunately, Paddlepaddle and PPOCRLabel work best with Python3.9. They do not work well with higher versions of Python, and you may encounter various issues. To make your life easier, start with 3.9. You also need to make sure you have Python installed directly on Windows. The easiest way to install Python nowadays is through the Microsoft store. Search for Python in the store and install the latest Python. Please note that we need Windows-native Python, not Python, under Windows Subsystem for Linux.

This will install both python.exe and pip.exe on your Windows, and you can use where command to check its path. To ensure they are installed successfully, use the following commands in Windows Command Prompt or Windows Powershell. It is always a good practice to upgrade pip before installing other packages.

where python
python --version
where pip
pip --version
pip install --upgrade pip

Step 2: Install Python Dependencies

There are a few pip dependencies that need to be installed.

pip install common
pip install dual
pip install tight
pip install data
pip install prox
pip install paddle

Step 3: Install PaddlePaddle

Follow the instructions from the official document to install paddlepaddle first. Install any missing pip packages if prompted.

python3 -m pip install paddlepaddle -i https://mirror.baidu.com/pypi/simple

Step 4: Install and Launch PPOCRLabel

pip install PPOCRLabel

# Select label mode and run
PPOCRLabel  # [Normal mode] for [detection + recognition] labeling
PPOCRLabel --kie True # [KIE mode] for [detection + recognition + keyword extraction] labeling

Now you are ready to label. Happy labeling!

PPOCRLabel

Conclusion

Installing PPOCRLabel on a Windows system is not straightforward, and you might encounter various problems. If you still cannot find a solution, comment below.

Leave a Reply

Your email address will not be published. Required fields are marked *