DetectAreasMode.TABLE

Contents
[ ]

When this mode is enabled, OCR engine detects tabular structures on an image and extracts text from cells. This areas detection mode is recommended when working with scanned spreadsheets, financial and accounting reports, invoices, and other tables.

However, this algorithm is inefficient when dealing with large amounts of text, such as pages from books and newspaper articles. Try DetectAreasMode.DOCUMENT instead.

Example

The following code sample demonstrates how to use this document areas detection algorithm:

Aspose.OCR.AsposeOcr recognitionEngine = new Aspose.OCR.AsposeOcr();
// Add an image to OcrInput object
Aspose.OCR.OcrInput input = new Aspose.OCR.OcrInput(Aspose.OCR.InputType.SingleImage);
input.Add("source.png");
// Set document areas detection mode
Aspose.OCR.RecognitionSettings recognitionSettings = new Aspose.OCR.RecognitionSettings();
recognitionSettings.DetectAreasMode = Aspose.OCR.DetectAreasMode.TABLE;
// Recognize image
List<Aspose.OCR.RecognitionResult> results = recognitionEngine.Recognize(input, recognitionSettings);
foreach(Aspose.OCR.RecognitionResult result in results)
{
	Console.WriteLine(result.RecognitionText);
}