OCR - Brain Dump – Oren Bochman’s Blog

I keep coming back to OCR.
OCR is a “Solved problem” but you can never get a decent OCR of your page….
Would be nice to have
1. some pdf image extraction code block in python.
2. some tools for generating text data in python - ideally with KL divergence to a corpus
3. some OCR building blocks in python
4. font manifold blocks in python + CLIPS capabilities
- make a manifold
- condition it
- a point at [font1-weight,font2-weight]
1. the power of off policy RL for this issue

Articles

Working of OCR articles by Susmith Reddy
- Pre-Processing in OCR covers Binarization, Skew Correction, Noise Removal, Thinning and Skeletonization
- Segmentation in OCR - Histogram Projection Method
Document Scanner app - Text Segmentation ^bibliography code by Arthur Flor some interesting algorithm.
Image Filters in Python code

PDF extraction

Extract text in a rectangle from pdf - Python
Pdftocairo man page

Font Manifolds

Learning a Manifold of Fonts Supplemental Material

Chat With AI:

Chat GPT

Citation

BibTeX citation:

@online{bochman2024,
  author = {Bochman, Oren},
  title = {OCR - {Brain} {Dump}},
  date = {2024-02-25},
  url = {https://orenbochman.github.io/posts/2024/2024-02-28-ocr/ocr.html},
  langid = {en}
}

For attribution, please cite this work as:

Bochman, Oren. 2024. “OCR - Brain Dump.” February 25, 2024. https://orenbochman.github.io/posts/2024/2024-02-28-ocr/ocr.html.