2021-04-03 22:34:49 +08:00
2021-04-03 22:34:49 +08:00
2021-01-17 22:13:44 +08:00
2021-02-19 11:17:14 +08:00
2021-01-17 22:13:44 +08:00
2021-01-17 22:13:44 +08:00
2021-02-06 12:06:40 +08:00
2021-02-07 07:35:48 +08:00
2021-01-17 22:13:44 +08:00
2021-01-23 20:20:44 +08:00
2021-03-30 14:48:20 +08:00
2021-04-01 14:06:53 +08:00
2021-04-03 21:45:17 +08:00
2021-02-06 12:06:40 +08:00
2021-03-27 10:35:31 +08:00
2021-01-17 22:13:44 +08:00
2021-01-24 16:42:08 +08:00
2021-03-27 21:58:15 +08:00
2021-03-01 14:30:33 +08:00
2021-02-08 02:57:09 +08:00
2021-04-03 22:14:16 +08:00

完全开源免费并支持离线部署的多平台多语言OCR SDK

中文广告: 欢迎加入我们的QQ群下载模型及测试程序,qq群号:887298230

欢迎Star 欢迎PR

SDK 编译状态

平台 编译状态 提供状态
Windows x86/x64 CMake-windows-x86-x64 右侧下载
Linux x64 CMake-linux 暂不提供,自行编译

常见问题

FAQ

重要说明

新模型已经完全兼容ONNXRuntime 1.7 或更高版本。 特别感谢:@Channingss

新版onnxruntime比1.6.0 性能提升40%以上。

在线演示(online demo): http://rapidocr.51pda.cn:9003/

演示模型组合说明: server det+mobile cls+mobile rec

中文详细说明,请访问: 中文版 Chinese Name: 捷智OCR

本仓库包括

  • 纯c/c++ API 接口,方便移植所有平台 (网盘提供预编译SDK)
  • .NET测试程序  演示在.NET程序中的使用方法
  • Android 测试程序
  • Java 测试程序
  • python 测试程序
  • WEB 演示 基于web的演示代码
  • IOS演示 招贡献者
  • 依据python版本重写C++推理代码,以提升推理效果,并增加对git/tga/webp 格式图片的支持

下载转换好的多语言模型:

Download OCR models

https://github.com/znsoftm/RapidOCR/releases/download/V1.0/rapid-model.tgz

---------------- For English Users ----------------------

License: Apache License 2.0

Note:

For PC (political correctness ), we changed our name from BaiPiaoOCR to RapidOCR

Special thanks to Channingss@baidu (the owner of project Paddle2Onnx)

Directory structure

Click to see 目录结构

Copyright announcement:

If you use or reference code or code snippet from the repository, please add our url

https://github.com/RapidOCR/RapidOCR in your prodouct derived from the repository.

Author:

benjaminwan, znsoftm

All contributors is in the contributor list on the right side of this page.

Description

BaiPiaoOCR means an OCR Engine who is fair-skinned, rich and beautiful from Baidu, it's based on PaddleOCR & OnnxRuntime.

Note: The project is derived from https://github.com/PaddlePaddle/PaddleOCR

We create the project to transfer PaddleOCR's model into the version of Onnx model to deploy the model on almost every devices, such as x86 PC, Android & IOS.

We are using Paddle2ONNX (https://github.com/PaddlePaddle/Paddle2ONNX) to transform it to Onnx format.

You can visualize the model by https://github.com/lutzroeder/netron/

Welcome to our Group on QQ: 887298230 or visit: https://jq.qq.com/?_wv=1027&k=P9b3olx6 to join us.

Download OCR models

Linkage :https://github.com/znsoftm/RapidOCR/releases/download/V1.0/rapid-model.tgz

Demonstration with C++/JVM

avatar

Demonstration with .Net

avatar

S
Description
RapidOCR - lightweight ONNX OCR
Readme Apache-2.0
74 MiB
0 Stars 1 Watchers 0 Forks
Languages
Python 81.3%
Jupyter Notebook 18.6%
Makefile 0.1%