RapidOCRSharpOnnx

🚀a high performance, cross-platform PaddleOCR C# inference library base on OpenCV and ONNX Runtime. Referring to the RapidOCR project, it is a python version of the C# implementation with a redesigned and optimized architecture.

Features

Supported Languages PP-OCRv6 built on the newly designed PPLCNetV4 unified backbone, a single model supports 50 languages including Chinese, English, Japanese, and 46 Latin-script languages., please refer to the documentation: PP-OCRv6 Introduction.
Execution Provider CPU, CUDA / TensorRT, OpenVINO, CoreML, DirectML
Batch processing images Preprocess and Inference are executed asynchronously with Producer/Consumer pattern
High Performance Inference Memory reuse, GPU Inference with I/O Binding
Image Processing OpenCvSharp4
Draw Result Image SkiaSharp
Inference Engine ONNX Runtime is a cross-platform inference and training machine-learning accelerator.
PP-OCR Versions Includes support for: PP-OCRv6, PP-OCRv5, PP-OCRv4

Example Images:

OCR Demo

Usage

1. Export model to ONNX format:

For convert the pre-trained PP-OCR model to ONNX format, please refer to the the documentation: Obtaining ONNX Models, or download from rapid-ocr Model List.

Official download of PP-OCRv6 onnx format

huggingface download

modelscope download

2. Load the ONNX model with C#:

Install Nuget packages RapidOCRSharpOnnx, OnnxRuntime, OpenCvSharp4.runtime

CPU Inference

dotnet add package RapidOCRSharpOnnx
dotnet add package OpenCvSharp4.runtime.win
dotnet add package Microsoft.ML.OnnxRuntime

using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderCPU(new OcrConfig(detectPath, recPath, LangRec.CH, OCRVersion.PPOCRV5, clsMobilePath)));

CoreML Inference

dotnet add package RapidOCRSharpOnnx
dotnet add package OpenCvSharp4.runtime.osx.10.15-x64
dotnet add package Microsoft.ML.OnnxRuntime

using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderCoreML(new OcrConfig(detectPath, recPath, LangRec.CH, OCRVersion.PPOCRV5, clsMobilePath)));

CUDA/TensorRT Inference

dotnet add package RapidOCRSharpOnnx
dotnet add package OpenCvSharp4.runtime.win
dotnet add package Microsoft.ML.OnnxRuntime.Gpu.Windows

using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderCUDA(new OcrConfig(detectPath, recogPath, LangRec.CH, OCRVersion.PPOCRV5, clsPath), deviceId));
using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderTensorRT(new OcrConfig(detectPath, recogPath, LangRec.CH, OCRVersion.PPOCRV5, clsPath), deviceId));

DirectML Inference

dotnet add package RapidOCRSharpOnnx
dotnet add package OpenCvSharp4.runtime.win
dotnet add package Microsoft.ML.OnnxRuntime.DirectML

using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderDirectML(new OcrConfig(detectPath, recogPath, LangRec.EN, OCRVersion.PPOCRV5, clsPath), deviceId));

OpenVINO Inference

dotnet add package RapidOCRSharpOnnx
dotnet add package OpenCvSharp4.runtime.win
dotnet add package Intel.ML.OnnxRuntime.OpenVino

using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderOpenVINO(new OcrConfig(detectPath, recogPath, LangRec.EN, OCRVersion.PPOCRV5, clsPath), IntelDeviceType.NPU));

Use the following C# code to load the model and run basic prediction

using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderDirectML(new OcrConfig(detectPath, recogPath, LangRec.EN, OCRVersion.PPOCRV5, clsPath), _deviceId));
string savePath = $"res_{Path.GetFileName(imgPath)}";
var result = ocr.RecognizeText(imgPath, savePath);
Console.WriteLine($"result: {result.ToString()}");

Batch processing images

 using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderDirectML(new OcrConfig(detectPath, recogPath, LangRec.CH, OCRVersion.PPOCRV5, clsPath), _deviceId));
 var list = Directory.GetFiles(@"C:\code\model\OCRTestImages");
 Stopwatch sw = new Stopwatch();
 sw.Start();
 var resPath = ocr.BatchParallelAsync(list.ToList(), saveDir, receiveAction: ReceiveResult);
 sw.Stop();
 Console.WriteLine($"BatchAsync Time: {sw.ElapsedMilliseconds} ms");

private static void ReceiveResult(OcrBatchResult batchResult)
{
    Console.WriteLine(batchResult.ToString());
    Console.WriteLine("------------------------------------------------------------");
}

Batch processing images foreach api

using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderDirectML(new OcrConfig(detectPath, recogPath, LangRec.CH, OCRVersion.PPOCRV5, clsPath), _deviceId));
var list = Directory.GetFiles(@"D:\code\model\OCRTestImages");
var resPaths = ocr.BatchForeachAsync(list.ToList(), @"D:\code\model\OCRTestImagesResults");

await foreach (var item in resPaths)
{
    Console.WriteLine(item.TextBlocks);
}

Custom font to draw result image

private static void TestImage()
{
     string imgPath = @"E:\Hp\ai-image\ADFtools\headerText.png";
     string detectPath = @"D:\code\RapidOCRSharpOnnx\RapidOCRSharpOnnx.TestCommon\Models\ch_PP-OCRv5_det_mobile.onnx";
     string recogPath = @"D:\code\RapidOCRSharpOnnx\RapidOCRSharpOnnx.TestCommon\Models\ch_PP-OCRv5_rec_mobile.onnx";
     string clsPath = @"D:\code\RapidOCRSharpOnnx\RapidOCRSharpOnnx.TestCommon\Models\ch_PP-LCNet_x0_25_textline_ori_cls_mobile.onnx";

     string font = @"C:\Windows\Fonts\msyh.ttc";
     using RapidOCRSharp ocr = new RapidOCRSharp(new ExecutionProviderDirectML(new OcrConfig(detectPath, recogPath, font, OCRVersion.PPOCRV5, clsPath), _deviceId));

     string savePath = $"res_{Path.GetFileName(imgPath)}";
     var result = ocr.RecognizeText(imgPath, savePath);
     Console.WriteLine($"result: {result.ToString()}");
}

Performance Test

OCR library	Version	language	Inference Engine
PaddleSharp	3.0.1	Paddle Inference C API .NET binding	Sdcb.PaddleInference
PaddleOCR	3.5.0	python	paddlepaddle
RapidOCR	3.8.1	python	openvino
RapidOCRSharpOnnx	1.0.0	C#	Intel.ML.OnnxRuntime.OpenVino

Performance Test PC

Hardware	Summary
Windows	Windows 11 Pro OS Version 25H2
CPU	Intel Core Ultra 9 285k 3.7GHz
RAM	DDR5 128GB speed 4400MT/s
Storage	SSD 2TB

Performance Test Data

Images: 60 images (image size: 1180x92)

PP-OCR Model: ch_PP-OCRv5_det_mobile, ch_PP-OCRv5_rec_mobile, ch_PP-LCNet_x0_25_textline_ori_cls_mobile

PaddleSharp test result

CPU inference time: 48.1769278s