OCR Result improvement

I am using tesseract dll with .NET application. In this where there are data in tables and and lines coming in image the text extracted are not accurate. The accuracy is too low. I tried all possible combinatins. How to improve this?

I tried PageSegMode and EngineMode all possible combination but the result are not appropriate. I am looking for improved accuracy. Also if any other open source tool available.

You can try leveraging the AutoPreprocess method to improve your image before you perform an OCR operation. In some cases, it would also be beneficial to investigate applying a suitable image processing technique to enhance the readability of the areas you want recognized.
Here is an example you can build on as well as documentation which shows the techniques you can use to improve your results.

static void OCRPreProcessing() 
{ 
   string tifFileName = @"C:\LEADTOOLS22\Resources\Images\Clean.tif"; 
   string pdfFileName = @"C:\LEADTOOLS22\Resources\Images\Clean.pdf"; 
 
   using (IOcrEngine ocrEngine = OcrEngineManager.CreateEngine(OcrEngineType.LEAD)) 
   { 
      // Start the engine using default parameters  
      ocrEngine.Startup(null, null, null, @"C:\LEADTOOLS22\Bin\Common\OcrLEADRuntime"); 
 
      // Create an OCR document  
      using (IOcrDocument ocrDocument = ocrEngine.DocumentManager.CreateDocument()) 
      { 
         // Add this image to the document  
         IOcrPage ocrPage = ocrDocument.Pages.AddPage(tifFileName, null); 
 
         // Auto-preprocess it  
         ocrPage.AutoPreprocess(OcrAutoPreprocessPageCommand.Deskew, null); 
         ocrPage.AutoPreprocess(OcrAutoPreprocessPageCommand.Invert, null); 
         ocrPage.AutoPreprocess(OcrAutoPreprocessPageCommand.Rotate, null); 
 
         // Recognize it and save it as PDF  
         ocrPage.Recognize(null); 
         ocrDocument.Save(pdfFileName, DocumentFormat.Pdf, null); 
      } 
   } 
}

https://www.leadtools.com/help/sdk/tutorials/dotnet-fx-console-preprocess-an-image-for-ocr.html
https://www.leadtools.com/help/sdk/v22/dh/po/namespace.html