cnn_lstm_ctc_ocr_for_ICPR
Python
Python
共35个文件
jpg: 11
py: 10
pyc: 4
tfrecord: 3
gitignore: 1
AUTHOR: 1
LICENSE: 1
Makefile: 1
md: 1
data/model/checkpoint: 1
Forked from weinman/cnn_lstm_ctc_ocr for the ICPR MTWI 2018 challenge 1
Forked from weinman/cnn_lstm_ctc_ocr for the ICPR MTWI 2018 challenge 1
Introduction
This is a repository forked from weinman/cnn_lstm_ctc_ocr for the ICPR MTWI 2018 challenge1.
Origin Repository: weinman/cnn_lstm_ctc_ocr - A tensorflow implementation of end to end text recognition
Origin Author: weinman
Author: Feng zhang
Email: [email protected]
Contents
Transform
Models
Demo
Train
Test
Results
Transform
You should cut the data by yourself.Use the target_cut.py, you only need to change the data path to your own origin data.I use the dataset like ICPR MTWI 2018. Note,I use Affine changes the change the Oblique picture to rectangle.So I need the clockwise direction of target label. Use the getTxt.py to change the label's direction.
Next, use mjsynth-tfrecord.py to change your data into tfrecord.You can Find the way in [weinman/cnn_lstm_ctc_ocr/Makefile] You only need to change some paths.
Models
I use the new word_dictionary which consists of English, Chinese and number.I only upload a old pretrain model,it works badly. If you train it for one day with your data,it will work well.And,I add some data augmentation for the model.
You also can change the model to denseNet,it will work better.Do as follows in the train.py :
Import denseNet
%features,sequence_length = model.convnet_layers( image, width, mode)
features,sequence_length = zf_mod_denseNet2.Dense_net( image, width, mode)
I only share the origin Model which is trained on ICPR MTWI 2018 (train): [model_download].The password is 2h1z.
Some English data can find in [weinman]
Demo
Download models and copy it to data/model. Then,Run:
python validate.py your picture's path
eg.python validate.py E:/1.jpg
Train
I already upload MTWI 2018 dataYou can download and copy to data/model/.After that ,train the model for a day,you can get your own model.[ICPR MTWI 2018 data+extra Eng-data+Eng-data].The password is diyj.
When we make your data to tfrecord,You can train. I suggest you had better use your own data.
cd src
python train.py
Test
use also can test the tfrecord's accuracy.[usage]
put your data in src/data/val and do as follows:
cd src;
python test.py
Results
Here are some results on ICPR MTWI 2018:
Enjoy yourself
没有合适的资源?快使用搜索试试~ 我知道了~
温馨提示
该项目是基于CNN_LSTM_CTC的OCR识别,专为ICPR MTWI 2018挑战赛1优化。主要使用Python语言编写,共包含35个文件,其中包括11个jpg图像文件,10个Python源代码文件,4个Python字节码文件,3个tfrecord数据文件,以及一些配置文件和模型文件。此项目是从weinman/cnn_lstm_ctc_ocr项目Fork而来,针对ICPR MTWI 2018挑战赛1进行了专门的优化。
资源推荐
资源详情
资源评论
收起资源包目录
upload.zip (35个子文件)
Makefile 631B
src
denseNet.py 6KB
word_dict.py 16KB
mjsynth-tfrecord.py 8KB
target_cut.py 5KB
mjsynth.py 8KB
model.py 8KB
validate.py 5KB
getTxt.py 3KB
train.py 8KB
__pycache__
mjsynth.cpython-36.pyc 5KB
denseNet.cpython-36.pyc 5KB
model.cpython-36.pyc 5KB
word_dict.cpython-36.pyc 16KB
test.py 7KB
data
val
words-000.tfrecord 1.46MB
train
words-000.tfrecord 1.46MB
model
checkpoint 289B
test_image
5_5.jpg 950B
2.jpg 2KB
result.jpg 18KB
1_1.jpg 895B
1.jpg 1KB
2_2.jpg 3KB
5.jpg 2KB
3.jpg 2KB
3_3.jpg 1KB
4.jpg 2KB
4_4.jpg 2KB
test
words-000.tfrecord 1.59MB
LICENSE 34KB
AUTHOR 69B
.gitignore 1KB
result
rank_42.png 15KB
readme.txt 2KB
共 35 条
- 1
资源评论
沐知全栈开发
- 粉丝: 4767
- 资源: 3372
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- 520马上到了两款动态爱心表白HTML代码(附源码)李峋同款爱心,快送给你爱的她或(他)吧12 情侣纪念日代码.zip
- 电子设计竞赛的单相不间断电源设计
- cutcamera1715961370938.png
- 基于MATLAB的图像处理课程设计报告.doc
- tensorflow-gpu-2.6.0-cp38-cp38-manylinux2010-x86-64.whl
- mmexport1715960553858.png
- tensorflow-gpu-2.6.0-cp37-cp37m-manylinux2010-x86-64.whl
- 通过 .NET 应用程序中的源代码查找 SQL 注入
- 电子设计竞赛2007年B题 无线识别装置.doc
- Wox全局搜索工具,一款win下的全局搜索软件
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功