Skip to content

ibyte2011/end-to-end-for-chinese-plate-recognition

 
 

Repository files navigation

end-to-end-for-plate-recognition

Python 3.5(64bit) 安装说明 http://www.cnblogs.com/ibyte/p/5871941.html

多标签分类,端到端的中文车牌识别基于mxnet . 从xlvector的ocr代码修改,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。

##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 image image image image image image image image image image image ##识别样张 ![image](./recognize_samples/Screen Shot 2016-08-07 at 12.51.56 AM.png) ![image](./recognize_samples/Screen Shot 2016-08-07 at 12.53.41 AM.png) ![image](./recognize_samples/Screen Shot 2016-08-07 at 12.55.45 AM.png)

About

多标签分类,端到端的中文车牌识别基于mxnet .

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%