Browse our Products

Aspose.OCR for .NET 21.7 - Release Notes

This page contains release notes information for Aspose.OCR for .NET 21.7

GPU version: 21.7.0

All Features

Key	Summary	Category
OCRNET-378	Confidence rating of potential match	Enhancement
OCRNET-376	Add new method (or parameter) that indicates the table exists on the image	Enhancement
OCRNET-375	Add a method that returns a list of rectangles of all lines in the text	Enhancement
OCRNET-372	Make Aspose.OCR works with x64, AnyCPU, x86 architecture	Enhancement

Enhancements

added new parameters in the method GetRectangles: AreasType areasType and bool isDetectAreas
added new property in the RecognitionSettings class that indicates the table exists on the image: LinesFiltration
added new property in the RecognitionResult class: RecognitionCharactersList
made Aspose.OCR works with x64, AnyCPU, x86 architectures
used compressed OCR ML model and reducing the packet size

Public API and Backwards Incompatible Changes

New API

added method: List GetRectangles(MemoryStream image, AreasType areasType = AreasType.PARAGRAPHS, bool isDetectAreas = true)
added method: List GetRectangles(string fullPath, AreasType areasType = AreasType.PARAGRAPHS, bool isDetectAreas = true)
added enum AreasType
added property: List<char[]> RecognitionCharactersList { get; internal set; } to the RecognitionResult class
added property: bool LinesFiltration { internal get; set; } = false to the RecognitionSettings class

Removed APIs

List GetRectangles(MemoryStream image)
List GetRectangles(string fullPath)

Will be deprecated

Usage Example

using Aspose.OCR;

namespace ProgramOCR
{
    class Program
    {
        static void Main(string[] args)
        {
            // Get API
            AsposeOcr api = new AsposeOcr();

            // Create license
            License lic = new License();

            // Set license 
            lic.SetLicense("Aspose.Total.lic");

            // Get image for recognize
            string imgPath = "img.png";

             // Get bounding boxes of lines            
            MemoryStream ms = new MemoryStream();
            using (FileStream file = new FileStream(imgPath, FileMode.Open, FileAccess.Read))
                file.CopyTo(ms);
            List<Rectangle> result = api.GetRectangles(ms, AreasType.LINES, true);

            // Get a list of possible character variants            
            RecognitionResult result = api.RecognizeImage(imgPath, new RecognitionSettings());
            foreach (var item in result.RecognitionCharactersList)
            {
                Console.WriteLine(item[0]+": "+item[1] + " "+item[2] + " "+item[3] + " "+item[4]);
            }
        }
    }
}