OCRHelper.cs


using System;
using System.Collections.Generic;
using System.Drawing;
using System.Text.RegularExpressions;
using Tesseract;
using System.IO;
using System.Windows.Forms;

namespace WindowsFormsApp7
{
    public static class OCRHelper
    {
        public static Bitmap PreprocessImage(Bitmap original)
        {
            Bitmap gray = new Bitmap(original.Width, original.Height);
            using (Graphics g = Graphics.FromImage(gray))
            {
                var colorMatrix = new System.Drawing.Imaging.ColorMatrix(
                    new float[][] {
                        new float[] { .3f, .3f, .3f, 0, 0 },
                        new float[] { .59f, .59f, .59f, 0, 0 },
                        new float[] { .11f, .11f, .11f, 0, 0 },
                        new float[] { 0, 0, 0, 1, 0 },
                        new float[] { 0, 0, 0, 0, 1 }
                    });
                var attributes = new System.Drawing.Imaging.ImageAttributes();
                attributes.SetColorMatrix(colorMatrix);
                g.DrawImage(original, new Rectangle(0, 0, gray.Width, gray.Height),
                            0, 0, original.Width, original.Height, GraphicsUnit.Pixel, attributes);
            }

       
            for (int y = 0; y < gray.Height; y++)
            {
                for (int x = 0; x < gray.Width; x++)
                {
                    Color pixel = gray.GetPixel(x, y);
                    int brightness = (pixel.R + pixel.G + pixel.B) / 3;
                    gray.SetPixel(x, y, brightness < 130 ? Color.Black : Color.White);
                }
            }

            return gray;
        }

        public static string ExtractTextFromImage(Bitmap image)
        {
            try
            {
                Bitmap processed = PreprocessImage(image);
                string tessDataPath = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "tessdata");
                if (!Directory.Exists(tessDataPath) || !File.Exists(Path.Combine(tessDataPath, "eng.traineddata")))
                {
                    MessageBox.Show("❌ لم يتم العثور على tessdata أو eng.traineddata!", "OCR Error", MessageBoxButtons.OK, MessageBoxIcon.Error);
                    return "[OCR error: Missing tessdata or traineddata]";
                }

                using (var engine = new TesseractEngine(tessDataPath, "eng", EngineMode.Default))
                {
                    using (var pix = PixConverter.ToPix(processed))
                    {
                        using (var page = engine.Process(pix))
                        {
                            string result = page.GetText();

                        
                            if (string.IsNullOrWhiteSpace(result) || result.Trim().Length < 5)
                            {
                                image.RotateFlip(RotateFlipType.Rotate180FlipNone);
                                processed = PreprocessImage(image);

                                using (var pix2 = PixConverter.ToPix(processed))
                                using (var page2 = engine.Process(pix2))
                                {
                                    result = page2.GetText();
                                }
                            }

                            return result;
                        }
                    }
                }
            }
            catch (Exception ex)
            {
                Console.WriteLine($"OCR error during image processing: {ex.Message}");
                return $"[OCR error: {ex.Message}]";
            }
        }

        public static List<string> ExtractTIDsFromText(string text)
        {
            List<string> tids = new List<string>();

          
            var matches = Regex.Matches(text, @"TID[\s:]*([A-Z0-9\-]{6,15})|(?<!\d)(\d{8})(?!\d)", RegexOptions.IgnoreCase);

            foreach (Match match in matches)
            {
                if (match.Groups.Count > 1)
                {
                    tids.Add(match.Groups[1].Value);
                }
            }
            return tids;
        }
    }
}