Core API

Core API

This page covers the object lifecycle and exception behavior of the four core types: AsposeOcr, OcrInput, OcrOutput, and RecognitionResult.

OcrInput Lifecycle

OcrInput implements IDisposable and should be created in a using statement. Its constructor takes an InputType (InputType.SingleImage or InputType.PDF), and OcrInput.Add(string) queues one file path at a time — the same OcrInput instance can hold multiple files of that same input type.

using Aspose.OCR;

using var input = new OcrInput(InputType.SingleImage);
input.Add("page-1.jpg");
input.Add("page-2.png");

Recognizing and Reading OcrOutput

AsposeOcr.Recognize(OcrInput) returns an OcrOutput, which is a List<RecognitionResult> — indexable like any list, with one RecognitionResult per image (for InputType.SingleImage, one per added file) or per PDF page (for InputType.PDF).

var api = new AsposeOcr();
OcrOutput output = api.Recognize(input);
RecognitionResult first = output[0];

RecognitionResult and the Nested LinesResult Type

RecognitionResult.LinesResult is a nested class — reference it as RecognitionResult.LinesResult, not as a standalone type. Each RecognitionResult exposes:

• RecognitionText — the complete recognized text for that image or page, with lines joined by line breaks.
• RecognitionLinesResult — a List<RecognitionResult.LinesResult>, one entry per detected text line, each with TextInLine (the line’s text) and Line (a System.Drawing.Rectangle giving that line’s pixel coordinates in the source image).

RecognitionResult result = output[0];
Console.WriteLine(result.RecognitionText);

foreach (RecognitionResult.LinesResult line in result.RecognitionLinesResult)
{
    System.Drawing.Rectangle bounds = line.Line;
    Console.WriteLine($"x={bounds.X}, y={bounds.Y}: {line.TextInLine}");
}

Exceptions

AsposeOcr.Recognize(OcrInput) validates its input before running recognition:

These checks run before any model inference, so a bad input fails immediately rather than after a partial recognition pass.

Tips and Best Practices

• OcrOutput is empty (Count == 0) if OcrInput has no files added — check Count before indexing.
• RecognitionResult.LinesResult is nested inside RecognitionResult; the top-level type name LinesResult alone is not a resolvable type outside that nesting.

Common Issues

FAQ

Is RecognitionResult.LinesResult its own top-level class?

No — it is nested inside RecognitionResult. Reference it as RecognitionResult.LinesResult.

Does OcrOutput guarantee page order for PDFs?

Yes — for InputType.PDF, OcrOutput returns one RecognitionResult per page in source order.

What happens if I call Recognize() with an empty OcrInput?

It returns an OcrOutput with Count == 0 rather than throwing.

API Reference Summary

See Also

• Aspose.OCR — Enterprise Documentation