https://github.com/szad670401/end-to-end-for-chinese-plate-recognition

多标签分类,端到端的中文车牌识别基于mxnet, End-to-End Chinese plate recognition base on mxnet
https://github.com/szad670401/end-to-end-for-chinese-plate-recognition

Last synced: 4 days ago
JSON representation

多标签分类,端到端的中文车牌识别基于mxnet, End-to-End Chinese plate recognition base on mxnet

• Host: GitHub
• URL: https://github.com/szad670401/end-to-end-for-chinese-plate-recognition
• Owner: szad670401
• Created: 2016-08-07T13:34:20.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2019-01-08T09:43:23.000Z (almost 6 years ago)
• Last Synced: 2024-10-30T03:42:53.468Z (17 days ago)
• Language: Python
• Homepage:
• Size: 35.9 MB
• Stars: 1,112
• Watchers: 62
• Forks: 522
• Open Issues: 29
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-ocr - 多标签分类,端到端的中文车牌识别基于mxnet, End-to-End Chinese plate recognition base on mxnet

README

        
# end-to-end-for-plate-recognition

多标签分类,端到端的中文车牌识别基于mxnet .

从[xlvector的ocr代码](https://github.com/szad670401/learning-dl/tree/master/mxnet/ocr)修改，减少了参数，由于我没有显卡。单线程 9 samples/s 速度 ，用CPU在MBP上跑了50w张样本。识别率到了81%。不过还没有完全收敛。

## 训练好的模型

https://github.com/ibyte2011/end-to-end-for-chinese-plate-recognition

## 关于车牌识别

生成的车牌对于实际车牌并不是效果很好，在结合真实样本和GAN，训练了一个更好的模型，对真实车牌表现很好。

并实现了一整套车牌识别的系统命名为HyperLPR https://github.com/zeusees/HyperLPR

## 依赖:

 + Numpy

 + Mxnet

 + Opencv

 

## 生成的车牌样张

通过渲染车牌加上畸变、噪声、与自然环境结合生成车牌的样本。

 ![image](./recognize_samples/00.jpg)

  ![image](./recognize_samples/01.jpg)

   ![image](./recognize_samples/02.jpg)

    ![image](./recognize_samples/03.jpg)

     ![image](./recognize_samples/04.jpg)

        ![image](./recognize_samples/06.jpg)

    ![image](./recognize_samples/07.jpg)

     ![image](./recognize_samples/08.jpg)   ![image](./recognize_samples/02.jpg)

    ![image](./recognize_samples/09.jpg)

     ![image](./recognize_samples/10.jpg)

         ![image](./recognize_samples/11.jpg)

## 识别样张







 

## Author

  + Jack Yu

  + Xiao Xiao