status: Alfa
Optical recognition system for electronic symbols, focused on the recognition of electronic circuits for export in SPICE format
optional not for now
- Make an image recognition model for electronic component diagrams (IEC)
- Recognize the connections between electronic components, in nodes
- Recognize the location of the components and vectorize the connections
Recognize the values of the components as well as any useful label (text) in it.- Value entry for each recognized component
- Generate a format that is read by a circuit simulator (SPICE)
- Simulate
How to Use
create python env if necessary
py -m venv .
install dependencies
pip install install huggingface-hub datasets tensorflow==2.15.0 opencv-python
python imagetest.py [classif_model_path] [ssd_model_path] [classname ...] [imagepath]
If you have compatibility problems with tensorflow with respect to other libraries, consider installing libraries from around Nov 14, 2023.
The code can easily run in Google Colab, because it was worked from said environment