https://github.com/watersink/ocrsegment

a deep learning model for page layout analysis / segmentation.
https://github.com/watersink/ocrsegment

Last synced: 10 days ago
JSON representation

a deep learning model for page layout analysis / segmentation.

• Host: GitHub
• URL: https://github.com/watersink/ocrsegment
• Owner: watersink
• Created: 2018-07-31T09:10:12.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2019-11-04T02:53:58.000Z (about 5 years ago)
• Last Synced: 2024-08-02T11:14:26.700Z (3 months ago)
• Language: Python
• Size: 2.19 MB
• Stars: 100
• Watchers: 6
• Forks: 27
• Open Issues: 6
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# OCR Segmentation

a deep learning model for page layout analysis / segmentation.

## dependencies

tensorflow1.8

> 

python3

## dataset:

[uw3-framed-lines-degraded-000](https://storage.googleapis.com/tmb-ocr/uw3-framed-lines-degraded-000.tgz)

## make training labels

    python3 data_pre_process.py

## train

    python3 train_test.py

## test

    python3 segmentation.py

![image]( https://github.com/watersink/ocrsegment/blob/master/make_training_labels/W001.png)

![image]( https://github.com/watersink/ocrsegment/blob/master/make_training_labels/out.png)

![image]( https://github.com/watersink/ocrsegment/blob/master/lines/0.png)

## references

[Multi-Dimensional Recurrent Neural Networks](https://arxiv.org/abs/0705.2011)


[Robust_ Simple Page Segmentation Using Hybrid Convolutional MDLSTM Networks](https://github.com/wanghaisheng/awesome-ocr/files/2042377/Robust_.Simple.Page.Segmentation.Using.Hybrid.Convolutional.MDLSTM.Networks.pdf)


[https://github.com/NVlabs/ocroseg](https://github.com/NVlabs/ocroseg)


[https://github.com/philipperemy/tensorflow-multi-dimensional-lstm](https://github.com/philipperemy/tensorflow-multi-dimensional-lstm)