lqy123000 / image_ocr

Image OCR in Keras

image_ocr

Image OCR in Keras.

Introduction

This project is inspired by image_ocr.py in keras examples.

Some main changes are listed as follows:

• Generate characters in random rather than loading words from file.
• Change network architecture. BN is added after Conv2D and one BiGRU is removed, since the origin network is found hard to converge.
• Reduce epochs since BN accelerates training.
• The function "on_train_begin" seems working in parallel with "fit_generator", so built word list on constructor.
• Replace cairocffi with PIL. cairocffi relies on GTK which is not easy to install on windows.
• If font size is too large to paint, try smaller size rather than throw a exception immediately.
• Rotation range is tuned to avoid string exceeding canvas.
• Support saving "predict_model".

Results

Test 10240 random images.

• digits test acc: 99.76% for training 15 epochs
• English character test acc: 96.16% for training 16 epochs
• captcha test acc: 82.93% for training 6 epochs (fix length 4)
• captcha_cnn test acc: 97.06% for training 10 epochs (fix length 4)

References

• keras example "image_ocr.py"
• Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks
• https://github.com/matterport/Mask_RCNN