kamal-lal / gesture_controlled_wincc

WinCC Advanced RT controlled using hand gestures

Gesture Controlled WinCC

WinCC Advanced RT controlled using hand gestures.

Introduction:

WinCC Advanced is an industrial 'Human Machine Interface' (HMI) system from SIEMENS. You can read more about it in this page. This project aims on sending commands to WinCC Advanced RT using hand gestures.

Project Structure:

The gesture recognition logic is written in Python. It uses OpenCV for the image processing and Tensorflow for the machine learning.

The visualization is done with WinCC Advanced. The Python script communicates with WinCC using OPC-UA channel. WinCC has native OPC-UA channel driver. For the Python module, "opcua" package is used.

Usage:

The Python part comprises of three files.

• A module to create image dataset for training the Neural Network (1_create_dataset.py).
• A module to train the Neural Network model (2_train_model.py).
• A module that uses the trained model to detect gestures and send commands to WinCC (3_gesture_control.py).

A detailed description on usage of individual module is given in module docstring. Trained model "gest_recog_model.h5" is also uploaded.

The WinCC RT project just has one screen. It reads 6 tags over OPC-UA to complete the demo.

Demo Video

Please see this LinkedIn post for a video demo.