renxiangyuan/slowctpn

OCR-DETECTION-CTPN

CNN+LSTM (CTPN) for image text detection

example results

description

To run this repo:

1, python data_base_normalize.py       # to normalize the pre-normalized background images

2, python data_generator.py 0       # to generate validation data

3, python data_generator.py 1       # to generate training data

4, python script_detect.py       # to train and validate

By 1, the pre-normalized images will firstly be rescaled if not of size 800x600, then 800x600 rects will be cropped from the rescaled images. The 800x600 images will be stored in a newly-maked directory, ./images_base.

By 2 and 3, validation data and training data will be generated. These will be store in the newly-maked directories, ./data_valid and ./data_train, respectively.

By 4, the model will be trained and validated. The validation results will be stored in ./data_valid/results. The ckpt files will be stored in a newly-maked directory, ./model_detect.

detection model

The model is mainly based on the method described in the article:

Detecting Text in Natural Image with Connectionist Text Proposal Network

Zhi Tian, Weilin Huang, Tong He, Pan He, Yu Qiao

https://arxiv.org/abs/1609.03605