This model is forked from ppt_detecteor
I add the training part and use it for detecting text in Chinese ID

• python 2.7
• keras=1.2.2
• theano=0.9.0
• some python libraries which can be included in Anaconda

X:\Users\Ray>python train_demo.py

prompt:you need create a list first and two folders(data and label),the format is similar to pascal_voc
just like
data/1.jpg label/1.xml
***
***

X:\Users\Ray>python test_demo.py