public static void PerformOCR(string imageFile, string outputFile, string langCode, string pageSegMode, bool hocr)
{
IList <Image> imageList;
try
{
imageList = ImageIOHelper.GetImageList(new FileInfo(imageFile));
OCR <Image> ocrEngine = new OCRImages();
ocrEngine.PageSegMode = pageSegMode;
ocrEngine.Hocr = hocr;
string result = ocrEngine.RecognizeText(imageList, langCode);
// skip post-corrections if hocr output
if (!hocr)
{
// postprocess to correct common OCR errors
result = Processor.PostProcess(result, langCode);
// correct common errors caused by OCR
result = TextUtilities.CorrectOCRErrors(result);
// correct letter cases
result = TextUtilities.CorrectLetterCases(result);
}
using (StreamWriter sw = new StreamWriter(outputFile, false, new System.Text.UTF8Encoding()))
{
sw.Write(result);
}
}
finally
{
imageList = null;
}
}
private void backgroundWorkerOcr_DoWork(object sender, DoWorkEventArgs e)
{
// Get the BackgroundWorker that raised this event.
BackgroundWorker worker = sender as BackgroundWorker;
OCRImageEntity entity = (OCRImageEntity)e.Argument;
OCR <Image> ocrEngine = new OCRImages();
ocrEngine.PageSegMode = selectedPSM;
ocrEngine.Language = entity.Language;
// Assign the result of the computation to the Result property of the DoWorkEventArgs
// object. This is will be available to the RunWorkerCompleted eventhandler.
//e.Result = ocrEngine.RecognizeText(entity.ClonedImages, entity.Lang, entity.Rect, worker, e);
IList <Image> images = entity.ClonedImages;
for (int i = 0; i < images.Count; i++)
{
if (worker.CancellationPending)
{
e.Cancel = true;
break;
}
string result = ocrEngine.RecognizeText(((List <Image>)images).GetRange(i, 1), entity.Inputfilename, entity.Rect, worker, e);
worker.ReportProgress(i, result); // i is not really percentage
}
}
/// <summary>
/// Performs OCR for bulk/batch and console operations.
/// </summary>
/// <param name="imageFile">Image file</param>
/// <param name="outputFile">Output file without extension</param>
/// <param name="langCode">language code</param>
/// <param name="pageSegMode">page segmentation mode</param>
/// <param name="outputFormat">format of output file. Possible values: <code>text</code>, <code>text+</code> (with post-corrections), <code>hocr</code></param>
public static void PerformOCR(string imageFile, string outputFile, string langCode, string pageSegMode, string outputFormat)
{
IList <Image> imageList;
try
{
DirectoryInfo dir = Directory.GetParent(outputFile);
if (dir != null && !dir.Exists)
{
dir.Create();
}
bool postprocess = "text+" == outputFormat;
OCR <Image> ocrEngine = new OCRImages();
ocrEngine.PageSegMode = pageSegMode;
ocrEngine.Language = langCode;
ocrEngine.OutputFormat = outputFormat.Replace("+", string.Empty);
// convert PDF to TIFF
if (imageFile.ToLower().EndsWith(".pdf"))
{
imageFile = PdfUtilities.ConvertPdf2Tiff(imageFile);
}
imageList = ImageIOHelper.GetImageList(new FileInfo(imageFile));
string result = ocrEngine.RecognizeText(imageList, imageFile);
// post-corrections for text+ output
if (postprocess)
{
// postprocess to correct common OCR errors
result = Processor.PostProcess(result, langCode);
// correct common errors caused by OCR
result = TextUtilities.CorrectOCRErrors(result);
// correct letter cases
result = TextUtilities.CorrectLetterCases(result);
}
//if (outputFormat == "pdf") // not yet supported
//{
// byte[] bytes = null; // get the byte array
// File.WriteAllBytes(outputFile, bytes);
//}
//else
{
string filename = outputFile + "." + outputFormat.Replace("+", string.Empty).Replace("text", "txt").Replace("hocr", "html");
using (StreamWriter sw = new StreamWriter(filename, false, new System.Text.UTF8Encoding()))
{
sw.Write(result);
}
}
}
finally
{
imageList = null;
}
}
private void AutoOCR()
{
FileInfo imageFile;
try
{
imageFile = new FileInfo(queue.Dequeue());
}
catch
{
return;
}
this.statusForm.TextBox.BeginInvoke(new UpdateStatusEvent(this.WorkerUpdate), new Object[] { imageFile.FullName });
if (curLangCode == null)
{
this.statusForm.TextBox.BeginInvoke(new UpdateStatusEvent(this.WorkerUpdate), new Object[] { " ** " + Properties.Resources.selectLanguage + " **" });
//queue.Clear();
return;
}
IList <Image> imageList = ImageIOHelper.GetImageList(imageFile);
if (imageList == null)
{
this.statusForm.TextBox.BeginInvoke(new UpdateStatusEvent(this.WorkerUpdate), new Object[] { " ** " + Properties.Resources.Cannotprocess + imageFile.Name + " **" });
return;
}
try
{
OCR <Image> ocrEngine = new OCRImages();
ocrEngine.PSM = selectedPSM;
string result = ocrEngine.RecognizeText(imageList, curLangCode);
// postprocess to correct common OCR errors
result = Processor.PostProcess(result, curLangCode);
// correct common errors caused by OCR
result = TextUtilities.CorrectOCRErrors(result);
// correct letter cases
result = TextUtilities.CorrectLetterCases(result);
using (StreamWriter sw = new StreamWriter(Path.Combine(outputFolder, imageFile.Name + ".txt"), false, new System.Text.UTF8Encoding()))
{
sw.Write(result);
}
}
catch (Exception e)
{
// Sets the UI culture to the selected language.
Thread.CurrentThread.CurrentUICulture = new CultureInfo(selectedUILanguage);
this.statusForm.TextBox.BeginInvoke(new UpdateStatusEvent(this.WorkerUpdate), new Object[] { " ** " + Properties.Resources.Cannotprocess + imageFile.Name + " **" });
Console.WriteLine(e.StackTrace);
}
}
void PerformOCR(string[] args)
{
try
{
if (args[0] == "-?" || args[0] == "-help" || args.Length == 1 || args.Length == 3)
{
Console.WriteLine("Usage: vietocr imagefile outputfile [-l langcode]");
return;
}
FileInfo imageFile = new FileInfo(args[0]);
FileInfo outputFile = new FileInfo(args[1]);
if (!imageFile.Exists)
{
Console.WriteLine("Input file does not exist.");
return;
}
string curLangCode;
if (args.Length == 2)
{
curLangCode = "eng"; //default language
}
else
{
curLangCode = args[3];
}
IList <Image> imageList = ImageIOHelper.GetImageList(imageFile);
OCR <Image> ocrEngine = new OCRImages();
string result = ocrEngine.RecognizeText(imageList, curLangCode);
// postprocess to correct common OCR errors
result = Processor.PostProcess(result, curLangCode);
// correct common errors caused by OCR
result = TextUtilities.CorrectOCRErrors(result);
// correct letter cases
result = TextUtilities.CorrectLetterCases(result);
using (StreamWriter sw = new StreamWriter(outputFile.FullName + ".txt", false, new System.Text.UTF8Encoding()))
{
sw.Write(result);
}
}
catch (Exception e)
{
Console.WriteLine("Error: " + e.Message);
}
}
/// <summary>
/// Performs OCR for bulk/batch and console operations.
/// </summary>
/// <param name="imageFile">Image file</param>
/// <param name="outputFile">Output file without extension</param>
/// <param name="langCode">language code</param>
/// <param name="pageSegMode">page segmentation mode</param>
/// <param name="outputFormat">format of output file. Possible values: <code>text</code>, <code>text+</code> (with post-corrections), <code>hocr</code></param>
public static void PerformOCR(string imageFile, string outputFile, string langCode, string pageSegMode, string outputFormat)
{
DirectoryInfo dir = Directory.GetParent(outputFile);
if (dir != null && !dir.Exists)
{
dir.Create();
}
bool postprocess = "text+" == outputFormat;
OCR <Image> ocrEngine = new OCRImages();
ocrEngine.PageSegMode = pageSegMode;
ocrEngine.Language = langCode;
ocrEngine.OutputFormat = outputFormat.Replace("+", string.Empty);
ocrEngine.OutputFile = outputFile;
// convert PDF to TIFF
if (imageFile.ToLower().EndsWith(".pdf"))
{
imageFile = PdfUtilities.ConvertPdf2Tiff(imageFile);
}
ocrEngine.ProcessFile(imageFile);
// post-corrections for text+ output
if (postprocess)
{
string filename = outputFile + ".txt";
string result = File.ReadAllText(filename);
// postprocess to correct common OCR errors
result = Processor.PostProcess(result, langCode);
// correct letter cases
result = TextUtilities.CorrectLetterCases(result);
using (StreamWriter sw = new StreamWriter(filename, false, new System.Text.UTF8Encoding()))
{
sw.Write(result);
}
}
}
protected void setSegmentedRegions()
{
if (!this.segmentedRegionsToolStripMenuItem.Checked || imageList == null || this.toolStripBtnActualSize.Enabled)
{
pictureBox1.SegmentedRegions = null;
pictureBox1.Refresh();
return;
}
OCR <Image> ocrEngine = new OCRImages();
Dictionary <Color, List <Rectangle> > map = pictureBox1.SegmentedRegions;
if (map == null)
{
map = new Dictionary <Color, List <Rectangle> >();
}
Bitmap image = (Bitmap)imageList[imageIndex];
List <Rectangle> regions = new List <Rectangle>();
if (toolStripMenuItemBlock.Checked)
{
if (!map.ContainsKey(Color.Gray))
{
regions = ocrEngine.GetSegmentedRegions(image, PageIteratorLevel.Block);
map.Add(Color.Gray, regions);
}
}
else
{
map.Remove(Color.Gray);
}
if (toolStripMenuItemPara.Checked)
{
if (!map.ContainsKey(Color.Green))
{
regions = ocrEngine.GetSegmentedRegions(image, PageIteratorLevel.Para);
map.Add(Color.Green, regions);
}
}
else
{
map.Remove(Color.Green);
}
if (toolStripMenuItemTextLine.Checked)
{
if (!map.ContainsKey(Color.Red))
{
regions = ocrEngine.GetSegmentedRegions(image, PageIteratorLevel.TextLine);
map.Add(Color.Red, regions);
}
}
else
{
map.Remove(Color.Red);
}
if (toolStripMenuItemWord.Checked)
{
if (!map.ContainsKey(Color.Blue))
{
regions = ocrEngine.GetSegmentedRegions(image, PageIteratorLevel.Word);
map.Add(Color.Blue, regions);
}
}
else
{
map.Remove(Color.Blue);
}
if (toolStripMenuItemSymbol.Checked)
{
if (!map.ContainsKey(Color.Magenta))
{
regions = ocrEngine.GetSegmentedRegions(image, PageIteratorLevel.Symbol);
map.Add(Color.Magenta, regions);
}
}
else
{
map.Remove(Color.Magenta);
}
pictureBox1.SegmentedRegions = map;
pictureBox1.Refresh();
//pictureBox1.Update();
}
void PerformOCR(string[] args)
{
try
{
if (args[0] == "-?" || args[0] == "-help" || args.Length == 1 || args.Length == 3 || args.Length == 5)
{
Console.WriteLine("Usage: vietocr imagefile outputfile [-l lang] [-psm pagesegmode]");
return;
}
FileInfo imageFile = new FileInfo(args[0]);
FileInfo outputFile = new FileInfo(args[1]);
if (!imageFile.Exists)
{
Console.WriteLine("Input file does not exist.");
return;
}
string curLangCode = "eng"; //default language
string psm = "3"; // or alternatively, "PSM_AUTO"; // 3 - Fully automatic page segmentation, but no OSD (default)
if (args.Length == 4)
{
if (args[2].Equals("-l"))
{
curLangCode = args[3];
}
else if (args[2].Equals("-psm"))
{
psm = args[3];
}
}
else if (args.Length == 6)
{
curLangCode = args[3];
psm = args[5];
try
{
Int16.Parse(psm);
}
catch
{
Console.WriteLine("Invalid input value.");
return;
}
}
IList <Image> imageList = ImageIOHelper.GetImageList(imageFile);
OCR <Image> ocrEngine = new OCRImages();
ocrEngine.PSM = psm;
string result = ocrEngine.RecognizeText(imageList, curLangCode);
// postprocess to correct common OCR errors
result = Processor.PostProcess(result, curLangCode);
// correct common errors caused by OCR
result = TextUtilities.CorrectOCRErrors(result);
// correct letter cases
result = TextUtilities.CorrectLetterCases(result);
using (StreamWriter sw = new StreamWriter(outputFile.FullName + ".txt", false, new System.Text.UTF8Encoding()))
{
sw.Write(result);
}
}
catch (Exception e)
{
Console.WriteLine("Error: " + e.Message);
}
}
