private void InitOcr(String path, String lang, OcrEngineMode mode)
{
try
{
if (_ocr != null)
{
_ocr.Dispose();
_ocr = null;
}
if (String.IsNullOrEmpty(path))
{
path = ".";
}
TesseractDownloadLangFile(path, lang);
TesseractDownloadLangFile(path, "osd"); //script orientation detection
String pathFinal = path.Length == 0 ||
path.Substring(path.Length - 1, 1).Equals(Path.DirectorySeparatorChar.ToString())
? path
: String.Format("{0}{1}", path, System.IO.Path.DirectorySeparatorChar);
_ocr = new Tesseract(pathFinal, lang, mode);
}
catch (System.Net.WebException e)
{
_ocr = null;
throw new Exception("Unable to download tesseract lang file. Please check internet connection.", e);
}
catch (Exception e)
{
_ocr = null;
throw e;
}
}
private void InitOcr(String path, String lang, OcrEngineMode mode)
{
try
{
if (_ocr != null)
{
_ocr.Dispose();
_ocr = null;
}
if (String.IsNullOrEmpty(path))
{
path = Tesseract.DefaultTesseractDirectory;
}
TesseractDownloadLangFile(path, lang);
TesseractDownloadLangFile(path, "osd"); //script orientation detection
_ocr = new Tesseract(path, lang, mode);
}
catch (System.Net.WebException e)
{
_ocr = null;
throw new Exception("Unable to download tesseract lang file. Please check internet connection.", e);
}
catch (Exception e)
{
_ocr = null;
throw e;
}
}
//"", "eng", OcrEngineMode.TesseractLstmCombined
public void Init(string path, string lang, OcrEngineMode mode)
{
if (_ocr != null)
{
_ocr.Dispose();
}
_ocr = new Tesseract(path, lang, mode);
}
private void InitializeOCR(String dataPath, String lang, OcrEngineMode mode)
{
try
{
if (_ocr != null)
{
_ocr.Dispose();
}
_ocr = new Tesseract(dataPath, lang, mode);
}
catch (Exception e)
{
_ocr = null;
throw e;
}
}
/// <summary>
/// Clear and reset the model. Required Init function to be called again before calling ProcessAndRender.
/// </summary>
public void Clear()
{
if (_ocr != null)
{
_ocr.Dispose();
_ocr = null;
}
}
public void SaveTicket()
{
try
{
var blockBlob = blobContainer.GetBlockBlobReference(ticket.PhotoName);
var photo = ConvertPhotoToBlackWhite(blockBlob);
if (photo == null)
{
return;
}
var text = ExtractTextFromPhoto(photo);
if (text == null)
{
return;
}
var lines = text.Split(new string[] { Environment.NewLine }, StringSplitOptions.RemoveEmptyEntries);
var lotteryNumberTable = CreateLotteryNumberTable();
foreach (string line in lines)
{
var lotteryNumber = ParseLotteryNumber(line);
if (!string.IsNullOrEmpty(lotteryNumber))
{
DataRow row = lotteryNumberTable.NewRow();
row["MatchNumbers"] = lotteryNumber.Substring(0, 10);
row["FinalNumbers"] = lotteryNumber.Substring(10, 2);
lotteryNumberTable.Rows.Add(row);
}
}
DataAccess.SaveTicket(ticket, lotteryNumberTable);
}
finally
{
if (OCR != null)
{
OCR.Dispose();
OCR = null;
}
}
}
public object Any(OCRRequest request)
{
OCRResponse myResponse = new OCRResponse();
if (request.Base64UploadFile == null || string.IsNullOrWhiteSpace(request.Base64UploadFile.ToString()))
{
myResponse.jsonResponse = "There was no file attached or no file name was specified.";
return(myResponse);
}
try
{
// get the image data and save to local storage
string base64ImgStr = request.Base64UploadFile.ToString().Split(',')[1];
byte[] tmpBytes = Convert.FromBase64String(base64ImgStr);
Image image = (Bitmap)((new ImageConverter()).ConvertFrom(tmpBytes));
// now do the OCR junk here
var ocr = new Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-.,/");
ocr.Init(@"C:\source\innovation\Content\tessdata", "eng", false);
var resizedImage = (Bitmap)Resize(image, (3000), (3000), false);
resizedImage.SetResolution(300, 300);
var blackAndWhiteImage = BlackAndWhite(resizedImage, new Rectangle(0, 0, resizedImage.Width, resizedImage.Height));
var result = ocr.DoOCR(blackAndWhiteImage, Rectangle.Empty);
OCRRawDataModel dataItems = new OCRRawDataModel();
dataItems.DataList = new List <OCRRawDataModel.RawDataItem>();
foreach (Word word in result)
{
var item = new OCRRawDataModel.RawDataItem();
item.Value = word.Text;
item.Confidence = (int)word.Confidence;
item.LineIndex = word.LineIndex;
dataItems.DataList.Add(item);
}
var mapper = new IdentificationCardMapper();
var mappedObjects = mapper.MapDriversLicenseData(dataItems);
myResponse.jsonResponse = JsonConvert.SerializeObject(mappedObjects);
// cleanup
image.Dispose();
resizedImage.Dispose();
ocr.Dispose();
}
catch (Exception ex)
{
// i don't know what to do here...
myResponse.jsonResponse = ex.Message;
}
return(myResponse);
}
public String readScreen(Bitmap bmp)
{
Tesseract tesseract = new Tesseract(@"C:\lang", "rus", OcrEngineMode.TesseractLstmCombined);
tesseract.SetImage(bmp.ToImage <Bgr, byte>());
tesseract.Recognize();
String text = tesseract.GetUTF8Text();
tesseract.Dispose(); //Очистка памяти
return(text);
}
private void loadLanguageToolStripMenuItem_Click(object sender, EventArgs e)
{
if (openLanguageFileDialog.ShowDialog() == System.Windows.Forms.DialogResult.OK)
{
_ocr.Dispose();
string path = Path.GetDirectoryName(openLanguageFileDialog.FileName);
string lang = Path.GetFileNameWithoutExtension(openLanguageFileDialog.FileName).Split('.')[0];
_ocr = new Tesseract(path, lang, OcrEngineMode.Default);
languageNameLabel.Text = String.Format("{0} : tesseract", lang);
}
}
// grayscales the image and then runs the recognizer over it.
private string ProcessImage(Mat image)
{
Tesseract tesseract = new Tesseract(TessdataDirectory, "eng", OcrEngineMode.Default, WhiteList, true);
Image <Gray, Byte> processedImage = image.ToImage <Gray, Byte>();
tesseract.SetImage(processedImage);
tesseract.Recognize();
string detectedText = tesseract.GetUTF8Text();
tesseract.Dispose();
return(detectedText);
}
public static void Main()
{
Stream stream = Assembly.GetExecutingAssembly().GetManifestResourceStream("TesserNet.Example.SkiaSharp.img.png");
SKBitmap image = SKBitmap.Decode(stream);
Tesseract tesseract = new Tesseract();
Console.WriteLine(tesseract.Read(image).Trim());
stream.Dispose();
image.Dispose();
tesseract.Dispose();
}
public static void Main()
{
Stream stream = Assembly.GetExecutingAssembly().GetManifestResourceStream("TesserNet.Example.System.Drawing.img.png");
Image image = Image.FromStream(stream);
Tesseract tesseract = new Tesseract();
Console.WriteLine(tesseract.Read(image).Trim());
stream.Dispose();
image.Dispose();
tesseract.Dispose();
}
public string TextRecognize(Image <Bgr, byte> imgSceneRect, Tesseract ocr)
{
string str;
//ocr.SetImage(FilterPlate(imgSceneRect));
ocr.SetImage(imgSceneRect);
ocr.Recognize();
str = ocr.GetUTF8Text();
ocr.Dispose();
return(str);
}
private void InitOCR(string dataPath, string lang, OcrEngineMode mode)
{
try
{
if (OCR != null)
{
OCR.Dispose();
OCR = null;
}
OCR = new Tesseract(dataPath, lang, mode);
}
catch (Exception e)
{
OCR = null;
Console.WriteLine("Failed to initialize tesseract OCR engine, error: " + e.Message);
}
}
private static void InitOcr(String path, String lang, OcrEngineMode mode)
{
try
{
if (_ocr != null)
{
_ocr.Dispose();
_ocr = null;
}
//_ocr = new Tesseract()
_ocr = new Tesseract(path, lang, mode, "1234567890");
//_ocr.SetVariable("tessedit_char_whitelist", "1234567890");
}
catch (Exception e)
{
_ocr = null;
}
}
//private static void TesseractDownloadLangFile(String folder, String lang)
//{
// String subfolderName = "tessdata";
// String folderName = System.IO.Path.Combine(folder, subfolderName);
// if (!System.IO.Directory.Exists(folderName))
// {
// System.IO.Directory.CreateDirectory(folderName);
// }
// String dest = System.IO.Path.Combine(folderName, String.Format("{0}.traineddata", lang));
// if (!System.IO.File.Exists(dest))
// using (System.Net.WebClient webclient = new System.Net.WebClient())
// {
// String source =
// String.Format("https://github.com/tesseract-ocr/tessdata/blob/4592b8d453889181e01982d22328b5846765eaad/{0}.traineddata?raw=true", lang);
// Console.WriteLine(String.Format("Downloading file from '{0}' to '{1}'", source, dest));
// webclient.DownloadFile(source, dest);
// Console.WriteLine(String.Format("Download completed"));
// }
//}
private void InitOcr(String path, String lang, OcrEngineMode mode)
{
try
{
if (_ocr != null)
{
_ocr.Dispose();
_ocr = null;
}
_ocr = new Tesseract(path, lang, mode);
}
catch (Exception e)
{
_ocr = null;
MessageBox.Show(e.Message, "Failed to initialize tesseract OCR engine", MessageBoxButtons.OK);
}
}
private void buttonRecognize_Click(object sender, EventArgs e)
{
try
{
string data = "";
_ocr = new Tesseract(Path.GetFullPath("../../tessdata/"), _language, _mode);
Cursor.Current = Cursors.WaitCursor;
_ocr.SetImage(_img_gray);
_ocr.Recognize();
data = _ocr.GetUTF8Text();
textRecognize.Text = data;
_ocr?.Dispose();
IDCardClient client;
switch (_type)
{
case DocumentType.IDCardFrontSide:
client = IDCardParser.ReadIDCardFrontSide(data);
textResult.Text = client.ToStringFrontSide();
break;
case DocumentType.IDCardBackSide:
client = IDCardParser.ReadIDCardBackSide(data);
textResult.Text = client.ToString();
break;
case DocumentType.INN:
string clientInn = INNParser.Parse(data);
textResult.Text = $"INN = {clientInn}";
break;
case DocumentType.SimpleText:
textResult.Text = data;
break;
default:
break;
}
} catch (Exception ex)
{
MessageBox.Show($"{ex.Message}");
}
}
private NumberRecognition()
{
try
{
if (_ocr != null)
{
_ocr.Dispose();
_ocr = null;
}
_ocr = new Tesseract(".\\", "eng", OcrEngineMode.TesseractLstmCombined);
_ocr.SetVariable("tessedit_char_whitelist", "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ");
}
catch (System.Net.WebException e)
{
_ocr = null;
throw new Exception("Unable to download tesseract lang file. Please check internet connection.", e);
}
catch (Exception e)
{
_ocr = null;
throw e;
}
}
/// <summary>
/// Disposes all the necessary objects
/// </summary>
protected override void DisposeObject()
{
_tesseract.Dispose();
}
protected override void DisposeObject()
{
_ocr.Dispose();
}
public void Dispose()
{
_ocr.Clear();
_ocr.Dispose();
_ocr = null;
}
void PerfromOCV()
{
string str1 = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz +-/1234567890{}()_<>\"\\;:*#%$&";
string language = "eng";
ICapture capture = (ICapture)video;
System.Threading.Thread.Sleep(2000);
Mat ImageMat = capture.QueryFrame();
IImage image1 = (IImage)ImageMat; //"Argument0", "Image Object");
language = "eng"; //"Argument1", "OCR Language");
int integerVariable1 = 70; //"Argument2", "Upper Left Corner X coordinate of ROI");
int integerVariable2 = 70; //"Argument3", "Upper Left Corner Y coordinate of ROI");
int integerVariable3 = 70; //"Argument4", "ROI Width (Number of Pixels");
int integerVariable4 = 70; //"Argument5", "ROI Height (Number of Pixels");
bool booleanVariable = false; //"Argument6", "Display Debug Window");
Image <Bgr, byte> image2 = new Image <Bgr, byte>(image1.Bitmap);
image2.ROI = new Rectangle(integerVariable1, integerVariable2, integerVariable3, integerVariable4);
string str2 = @".\\tessdata";
if (!Directory.Exists(str2))
{
Directory.CreateDirectory(str2);
}
HelperMethods.LoadLanguage(str2, language);
HelperMethods.LoadLanguage(str2, "osd");
//HelperMethods.LoadLanguage(str2, language);
//string path = Path.GetFullPath(@"C:\Users\MarquezFr\source\repos\EmguOpenCVExample\EmguOpenCVExample\bin\Debug\tessdata\");
var tesseract = new Tesseract(Path.GetFullPath(str2.Length == 0 || str2.Substring(str2.Length - 1, 1).Equals(Path.DirectorySeparatorChar.ToString()) ? str2 : string.Format("{0}{1}", (object)str2, (object)Path.DirectorySeparatorChar)), language, OcrEngineMode.Default);
tesseract.Init(Path.GetFullPath(str2.Length == 0 || str2.Substring(str2.Length - 1, 1).Equals(Path.DirectorySeparatorChar.ToString()) ? str2 : string.Format("{0}{1}", (object)str2, (object)Path.DirectorySeparatorChar)), language, OcrEngineMode.Default);
//Tesseract tesseract = new Tesseract(path, language, OcrEngineMode.TesseractOnly);
tesseract.SetVariable("tessedit_char_whitelist", str1);
Image <Bgr, byte> image3 = new Image <Bgr, byte>(image2.Bitmap);
if (image1.NumberOfChannels == 1)
{
CvInvoke.CvtColor((IInputArray)image2, (IOutputArray)image3, ColorConversion.Gray2Bgr, 0);
}
else
{
//image3 = (Image<Bgr, byte>)image1.Clone(); //original
image3 = image1.Clone() as Image <Bgr, byte>;
}
//tesseract.SetImage((IInputArray)image3); original
tesseract.SetImage((IInputArray)image3);
if (tesseract.Recognize() != 0)
{
MessageBox.Show(string.Format("OpenCV_PerformOCR: OCR system failed to recognize image. OCR returned: {0}", (object)tesseract.Recognize()));
}
string utF8Text = tesseract.GetUTF8Text();
char[] separator = new char[3]
{
' ',
'\r',
'\n'
};
int length = utF8Text.Split(separator, StringSplitOptions.RemoveEmptyEntries).Length;
if (booleanVariable)
{
//ImageViewer imageViewer = new ImageViewer((IImage)image2);
//int num = (int)imageViewer.ShowDialog();
//imageViewer.Dispose();
}
image2.Dispose();
image3.Dispose();
tesseract.Dispose();
}
