Full, Multi-Language Supported C++ & ONNXRuntime Implementation of EasyOCR 基于C++和ONNXRuntime的EasyOCR完整移植(具备多语言支持能力)
You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
UnknownObject 1c65b30a64 fix repeat char bug 1 week ago
DLLUsageDemo change recognizer output format 2 months ago
EasyOCR-CPP fix repeat char bug 1 week ago
LibEasyOCR-CPP fix repeat char bug 1 week ago
.gitattributes 添加 .gitattributes、.gitignore、README.md 和 LICENSE.txt 2 months ago
.gitignore 添加 .gitattributes、.gitignore、README.md 和 LICENSE.txt 2 months ago
EasyOCR-CPP.sln add dll version and its usage 2 months ago
LICENSE.txt 添加 .gitattributes、.gitignore、README.md 和 LICENSE.txt 2 months ago
README.md change model name & write readme 2 months ago

README.md

EasyOCR-CPP

A C++ implementation of EasyOCR using ONNXRuntime.

Library Needed

  • OpenCV 4.10
  • ONNXRuntime 1.19.2
  • NVIDIA CUDA 11.7 (Optional)

C++ Version

C++17 or above.

Build

Using Visual Studio 2022 or later, open the EasyOCR-CPP.sln file and build the EasyOCR-CPP project.

Model Info

Detection Model

Detection model can handle all kinds of language that EasyOCR can recognize.

Recognition Model

Recognition model is different for each language. In this project, the English and English+Chinese model are provided.

Usage

See EasyOCR-CPP/EasyOCR-CPP.cpp for an example usage.

Todo

Provide A modified version of EasyOCR that can export any recognition model to ONNX format easily.

Other Informations

EasyOCR Library: https://github.com/JaidedAI/EasyOCR ONNX Model Export Guide: https://github.com/JaidedAI/EasyOCR/issues/746