A project for parsing race photos and identifying bib numbers.
bib-tagger will read in a directory of photos and attempt to identify each visible bib number.
- face detection, body detection
- bib detection
- SWT on bib
- OCR on SWT results
Probably could stand to clean these up a bit, do we need all of these? *Scipy *Numpy *pytesseract **tesseract-ocr *Pillow
Main.py will run through all of the file in a directory labeled 'photos'. modify tests\test_bibtagger.py to test the entire pipeline on a single file.
- Fork it!
- Create your feature branch:
git checkout -b my-new-feature - Commit your changes:
git commit -am 'Add some feature' - Push to the branch:
git push origin my-new-feature - Submit a pull request :D
- Refining SWT & OCR based on smaller problem set of bib detection
- Create results hash map
First version created by KateRita & Ryanr23
- Facial/Body Detection Reference: http://people.csail.mit.edu/talidekel/papers/RBNR.pdf
- SWT Reference: http://research.microsoft.com/pubs/149305/1509.pdf
- SWT Reference/Starting Point: https://github.com/mypetyak/StrokeWidthTransform/blob/master/swt.py
- OCR C++ Library: https://github.com/tesseract-ocr/tesseract
- OCR Python Wrapper: https://pypi.python.org/pypi/pytesseract