private void DefaultSettings()
{
engine.SetVariable("tessedit_char_whitelist", "0123456789"); // show only digits
engine.SetVariable("load_system_dawg", false); // disable dictionary values
engine.SetVariable("load_freq_dawg", false); // disable dictionary values
engine.DefaultPageSegMode = PageSegMode.SingleWord;
}
/// <summary>
/// 识别印刷票号
/// </summary>
/// <param name="imgs"></param>
/// <returns></returns>
private string OCRTicketNo(IList <Bitmap> imgs)
{
string res = "";
using (var engineLetter = new TesseractEngine(@"tessdata", "eng", EngineMode.TesseractOnly))
{
engineLetter.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ");
engineLetter.SetVariable("tessedit_unrej_any_wd", true);
engineLetter.DefaultPageSegMode = PageSegMode.SingleChar;
using (var page = engineLetter.Process(imgs[0], PageSegMode.SingleChar))
res += page.GetText().Substring(0, 1);
}
imgs.RemoveAt(0);
using (var engine = new TesseractEngine(@"tessdata", "eng", EngineMode.TesseractOnly))
{
engine.SetVariable("tessedit_char_whitelist", "1234567890");
engine.SetVariable("tessedit_unrej_any_wd", true);
engine.DefaultPageSegMode = PageSegMode.SingleChar;
foreach (Bitmap img in imgs)
{
using (var page = engine.Process(img, PageSegMode.SingleChar))
res += page.GetText().Substring(0, 1);
}
}
Console.WriteLine("OCR Result = " + res);
return(res);
}
public static string OCR(Bitmap b)
{
try
{
string res = string.Empty;
string path = $@"{Environment.CurrentDirectory}\tessdata\";
using (var engine = new TesseractEngine(path, "eng"))
{
string letters = "abcdefghijklmnopqrstuvwxyz";
string numbers = "0123456789";
engine.SetVariable("tessedit_char_whitelist", $"{numbers}{letters}{letters.ToUpper()}");
engine.SetVariable("tessedit_unrej_any_wd", true);
engine.SetVariable("tessedit_adapt_to_char_fragments", true);
engine.SetVariable("tessedit_redo_xheight", true);
engine.SetVariable("chop_enable", true);
Bitmap x = b.Clone(new Rectangle(0, 0, b.Width, b.Height), System.Drawing.Imaging.PixelFormat.Format24bppRgb);
using (var page = engine.Process(x, PageSegMode.SingleLine))
res = page.GetText().Replace(" ", "").Trim();
}
return(res);
}
catch (Exception ex)
{
////MessageBox.Show($"Erro: {ex.Message}");
return(null);
}
}
// 464/19516
static void Main(string[] args)
{
int hit, all;
hit = all = 0;
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default))
{
foreach (var file in Directory.EnumerateFiles(@"D:\dataset\easy\first try\read_and_segmented"))
{
string code = Path.GetFileNameWithoutExtension(file);
engine.DefaultPageSegMode = PageSegMode.SingleLine;
engine.SetVariable("load_system_dawg", "false");
engine.SetVariable("load_freq_dawg", "false");
engine.SetVariable("tessedit_char_whitelist", "-2346789BCDFGHJKMPQRTVWXY");
using (var img = Pix.LoadFromFile(file))
using (var page = engine.Process(img))
{
var text = page.GetText().Trim();
if (text == code)
{
++hit;
}
}
++all;
}
}
Console.WriteLine($"{hit}/{all}");
Console.ReadKey();
}
/// <summary>
/// Reads tessdata/configs/tess_configvars and SetVariable on Tesseract engine.
/// This only works for non-init parameters (@see <a href="https://code.google.com/p/tesseract-ocr/wiki/ControlParams">ControlParams</a>).
/// </summary>
/// <param name="engine"></param>
void ControlParameters(TesseractEngine engine)
{
string configsFilePath = Path.Combine(Datapath, "tessdata/configs/" + CONFIGVARS_FILE);
if (!File.Exists(configsFilePath))
{
return;
}
string[] lines = File.ReadAllLines(configsFilePath);
foreach (string line in lines)
{
if (!line.Trim().StartsWith("#"))
{
try
{
string[] keyValuePair = line.Trim().Split(new char[0], StringSplitOptions.RemoveEmptyEntries);
string value = keyValuePair[1];
if (value == "T" || value == "F")
{
engine.SetVariable(keyValuePair[0], value == "T" ? true : false);
}
else
{
engine.SetVariable(keyValuePair[0], keyValuePair[1]);
}
}
catch
{
//ignore and continue on
}
}
}
}
private void InitTesseract()
{
var tesseractData = Path.Combine(Path.GetDirectoryName(Assembly.GetExecutingAssembly().Location), "tessdata");
_englishOcr = new TesseractEngine(tesseractData, "eng", EngineMode.TesseractAndLstm);
_englishOcr.DefaultPageSegMode = PageSegMode.SingleLine;
_englishOcr.SetVariable("tessedit_char_whitelist", "qwertyuiopasdfghjklzxcvbnm QWERTYUIOPASDFGHJKLZXCVBNM");
_englishOcr.SetVariable("tessedit_zero_rejection", true);
_englishOcr.SetVariable("load_freq_dawg", false);
_englishOcr.SetVariable("load_system_dawg", false);
_englishPunctOcr = new TesseractEngine(tesseractData, "eng", EngineMode.TesseractAndLstm);
_englishPunctOcr.DefaultPageSegMode = PageSegMode.SingleLine;
_englishPunctOcr.SetVariable("tessedit_char_whitelist", "qwertyuiopasdfghjklzxcvbnm QWERTYUIOPASDFGHJKLZXCVBNM0123456789/*-+:,.%");
_englishPunctOcr.SetVariable("tessedit_zero_rejection", true);
_englishPunctOcr.SetVariable("load_freq_dawg", false);
_englishPunctOcr.SetVariable("load_system_dawg", false);
_numbersOcr = new TesseractEngine(tesseractData, "eng", EngineMode.TesseractAndLstm);
_numbersOcr.DefaultPageSegMode = PageSegMode.SingleLine;
_numbersOcr.SetVariable("tessedit_char_whitelist", "0123456789");
_numbersOcr.SetVariable("tessedit_zero_rejection", true);
_numbersOcr.SetVariable("load_freq_dawg", false);
_numbersOcr.SetVariable("load_system_dawg", false);
_numbersScaledOcr = new TesseractEngine(tesseractData, "eng", EngineMode.TesseractAndLstm);
_numbersScaledOcr.DefaultPageSegMode = PageSegMode.SingleLine;
_numbersScaledOcr.SetVariable("tessedit_char_whitelist", "0123456789.,KMB");
_numbersScaledOcr.SetVariable("tessedit_zero_rejection", true);
_numbersScaledOcr.SetVariable("load_freq_dawg", false);
_numbersScaledOcr.SetVariable("load_system_dawg", false);
Logger.Info("Tessract engines initialized");
}
public static void Run()
{
var bitmap = new Bitmap("E:\\dev\\venomsw\\images\\cropped_rune.png");
var scaled = ScaleBitmap(bitmap, bitmap.Width * 2, bitmap.Height * 2);
//TODO debug only
if (analyzer.ShouldGetRune(scaled))
{
return;
}
TesseractEngine engine = new TesseractEngine(@"E:\\dev\\venomsw\\venomsw\\tessdata", "eng",
EngineMode.Default, "venom");
engine.SetVariable("language_model_penalty_non_freq_dict_word", "1");
engine.SetVariable("language_model_penalty_non_dict_word", "1");
using (Page page = engine.Process(scaled, PageSegMode.SingleBlock))
{
Console.WriteLine(page.GetText());
}
scaled.Save("E:\\dev\\venomsw\\images\\cropped_rune3_nobg.png");
}
public SVM(string TrainedDataInputFile) //// Do training for all existing trained Data
{
_engine = new TesseractEngine(@"./tessdata3", "eng", EngineMode.TesseractAndCube);
_engine.SetVariable("tessedit_char_whitelist", "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ");
_engine.SetVariable("tessedit_char_blacklist", "¢§+~»~`!@#$%^&*()_+-={}[]|\\:\";\'<>?,./");
string[] TrainedData = Directory.GetFiles(TrainedDataInputFile, "*.png");
double[][] inputs = new double[TrainedData.Length][]; ///
double[] InputArray = new double[784];
int[] Outputs = new int[TrainedData.Length];
for (int i = 0; i < TrainedData.Length; i++)
{
string filename = Path.GetFileNameWithoutExtension(TrainedData[i]);
Bitmap TrainingImage = new Bitmap(TrainedData[i]);
string[] split = filename.Split('.');
for (int j = 0; j < 28; j++)
{
for (int k = 0; k < 28; k++)
{
if ((!TrainingImage.GetPixel(j, k).Name.Equals("ffffffff")))
{
InputArray[j * 28 + k] = 1;
}
else
{
InputArray[j * 28 + k] = 0;
}
}
}
inputs[i] = InputArray;
Outputs[i] = Convert.ToInt32(split[0]);
InputArray = new double[784];
}
IKernel kernel;
kernel = new Polynomial(2, 0);
ksvm = new MulticlassSupportVectorMachine(784, kernel, 2);
MulticlassSupportVectorLearning ml = new MulticlassSupportVectorLearning(ksvm, inputs, Outputs);
double complexity = 1; ///// set these three parameters Carefuly later
double epsilon = 0.001;
double tolerance = 0.2;
ml.Algorithm = (svm, classInputs, classOutputs, i, j) =>
{
var smo = new SequentialMinimalOptimization(svm, classInputs, classOutputs);
smo.Complexity = complexity; /// Cost parameter for SVM
smo.Epsilon = epsilon;
smo.Tolerance = tolerance;
return(smo);
};
// Train the machines. It should take a while.
double error = ml.Run();
}
public static string OCR(Bitmap imagem, string linguagem)
{
string texto = "";
using (TesseractEngine engine = new TesseractEngine(@"C:\GitHub\operacao-politica-supervisionada\OPS\temp\", linguagem, EngineMode.Default)) {
engine.SetVariable("tessedit_char_whitelist", "1234567890ABCDEFGHIJKLMNOPQRSTUVWXYZ");
engine.SetVariable("tessedit_unrej_any_wd", true);
engine.SetVariable("applybox_learn_chars_and_char_frags_mode", true);
engine.SetVariable("save_blob_choices", true);
string sobreposto = "";
int ultimo = 12;
using (Page page = engine.Process(imagem, PageSegMode.SingleLine)) {
using (ResultIterator ri = page.GetIterator()) {
do
{
string word = ri.GetText(PageIteratorLevel.Symbol);
Tesseract.Rect bb;
if (ri.TryGetBoundingBox(PageIteratorLevel.Symbol, out bb))
{
if ((bb.Width > 13) && (bb.Height > 15) && (word.Trim() != ""))
{
while (bb.X1 > ultimo + 14)
{
texto += Resolver(sobreposto);
sobreposto = "";
ultimo += 28;
}
//System.Web.HttpContext.Current.Response.Write(word + ": " + bb.X1 + "<br />\n");
if ((word != "Q") || (bb.Height <= 30))
{
sobreposto += word;
}
else
{
sobreposto += "O";
}
}
}
} while((ri.Next(PageIteratorLevel.Symbol)));
if (texto.Length < 6)
{
texto += Resolver(sobreposto);
while (texto.Length < 6)
{
texto += LetraAleatoria();
}
}
}
}
}
return(texto);
}
private static string OCRItalien(Bitmap b)
{
string res = "";
using (var engine = new TesseractEngine(null, "eng", EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist", "1234567890+-");
engine.SetVariable("tessedit_unrej_any_wd", true);
using (var page = engine.Process(b, PageSegMode.SingleLine))
res = page.GetText();
}
return res;
}
// Called for each strip, first cleans the image of noise and then performs OCR on the strip.
private string recognizeStripName(string fileName, string allowedChars = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ\'\"")
{
string noiseFreeImage = Path.Combine(Configuration.Instance.TempFolder, "temp.png");
// Make image white text on black background:
cleanImageAndSave(fileName, noiseFreeImage, isNotInWhiteRange);
// Use only specific characters:
m_ocrEngine.SetVariable("tessedit_char_whitelist", allowedChars);
string retVal = PerformOCR(noiseFreeImage, false);
File.Delete(noiseFreeImage);
return(retVal);
}
public static TesseractEngine GetTesseractEngine()
{
if (m_ocr == null)
{
m_ocr = new TesseractEngine("./tessdata", "eng", EngineMode.TesseractAndCube);
m_ocr.SetVariable("tessedit_char_whitelist", "01234567890");
m_ocr.SetVariable("tessedit_char_blacklist", "l");
return(m_ocr);
}
else
{
return(m_ocr);
}
}
public static string GetTextFromImage(string path)
{
var ocrtext = string.Empty;
Bitmap image = new Bitmap(path);
using (var engine = new TesseractEngine("tessdata", "eng+por", EngineMode.TesseractAndCube))
{
engine.SetVariable("tessedit_char_whitelist", "1234567890ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyzçÇãÃáàõéúí");
engine.SetVariable("tessedit_unrej_any_wd", true);
using (var page = engine.Process(image))
ocrtext = page.GetText();
}
return(ocrtext);
}
private string OCR(Bitmap b)
{
string res = "";
using (var engine = new TesseractEngine(@"tessdata", "spa", EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist", "1234567890abcdefghijklmnopqrstuvwxyz");
engine.SetVariable("tessedit_unrej_any_wd", true);
using (var page = engine.Process(b, PageSegMode.SingleBlock))
res = page.GetText();
}
return(res);
}
private List <String> GetNewObscureWords(Bitmap imageFile)
{
Bitmap extractMissedObscureWordsInImage = ExtractMissedObscureWordsInImage(imageFile);
TesseractEngine engine = new TesseractEngine(Program.TessDataDir, "eng", EngineMode.Default);
engine.SetVariable("tessedit_char_whitelist", "abcdefghijklmnopqrstuvwxyz");
engine.SetVariable("tessedit_char_blacklist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789!@#$%^&*()-_=\\|/?"); // no digit
Page page = engine.Process(extractMissedObscureWordsInImage, PageSegMode.SingleBlock);
String text = page.GetText();
return(text.Split(new [] { "\n" }, StringSplitOptions.RemoveEmptyEntries).ToList());
}
public static string OCR(Bitmap b)
{
string res = "";
using (var engine = new TesseractEngine(@"C:\RODRIGO\PROJETOS\ALTRAN\AeC - Framewors\AeC.Automacao.ImageProcess\AeC.ImageProcess\tessdata", "eng", EngineMode.TesseractOnly))
{
engine.SetVariable("tessedit_char_whitelist", "1234567890ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvxz");
engine.SetVariable("tessedit_unrej_any_wd", true);
using (var page = engine.Process(b, PageSegMode.SingleLine))
res = page.GetText();
}
return(res);
}
private string OCR(Bitmap b)
{
string res = "";
using (var engine = new TesseractEngine(@"tessdata", "eng", EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist", "1234567890ABCDEFGHIJKLMNOPQRSTUVWXYZ");
engine.SetVariable("tessedit_unrej_any_wd", true);
using (var page = engine.Process(b, PageSegMode.SingleLine))
res = page.GetText();
}
return(res);
}
private static void Initialize(ILogger _logger)
{
_logger.LogInformation("Initializing static Tesseract");
Tesseract = new TesseractEngine(@"./tessdata", "eng", EngineMode.TesseractAndLstm)
{
DefaultPageSegMode = PageSegMode.SingleBlock
};
Tesseract.SetVariable("tessedit_char_whitelist", ALLOWED_CHARACTERS);
Tesseract.SetVariable("user_patterns_file", @"./tessdata/cyber_patterns");
// \n (char or digit), \c (char), \d (digit), \p (punct), \a (lower), \A (upper), \* any number (\A\d and \d\A)
Tesseract.SetVariable("user_words_file", @"./tessdata/cyber_words"); // BD, 1C, E9, 55, 7A, 1F
Tesseract.SetVariable("load_system_dawg", false); // Don't load sys dictionary
Tesseract.SetVariable("load_freq_dawg", false); // Don't load word freeuence
}
private char[] RetrieveCharsByOcr(Bitmap data, string tessDataDir)
{
// ocr.SetVariable("tessedit_char_blacklist", "0123456789"); // no digit
// ocr.SetVariable("tessedit_char_whitelist", "abcdefghijklmnopqrstuvwxyz"); // yes letters
// ocr.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ");
TesseractEngine engine = new TesseractEngine(tessDataDir, "eng", EngineMode.Default);
engine.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ");
engine.SetVariable("tessedit_char_blacklist", "abcdefghijklmnopqrstuvwxyz0123456789!@#$%^&*()-_=\\|/?"); // no digit
Page page = engine.Process(data, PageSegMode.SingleWord);
//Console.WriteLine("{0} : {1}", page.GetMeanConfidence(), page.GetText());
return(page.GetText().Replace("\n", "").ToLower().ToCharArray());
}
private void Form1_Load(object sender, EventArgs e)
{
//Il caricamento dei file per tesseract viene fatto una volta solo al momento dell'apertura del programma
ocrengine = new TesseractEngine(@".\tessdata", "ita", EngineMode.Default);
ocrengine.DefaultPageSegMode = PageSegMode.SingleBlock;
ocrengine.SetVariable("tessedit_char_whitelist", "0123456789-_abcdfghmnpqrtuvzxyk");
}
public String DoOCR([FromForm] OcrModel request)
{
string name = request.Image.FileName;
var image = request.Image;
var imageStream = new MemoryStream();
if (image.Length > 0)
{
image.CopyTo(imageStream);
}
string tessPath = Path.Combine(trainedDataFolderName, "");
string result = "";
// TODO: Create one instance of engine and inject into app
using (var engine = new TesseractEngine(tessPath, request.DestinationLanguage, EngineMode.Default))
{
// whitelist numbers only
engine.SetVariable("tessedit_char_whitelist", "0123456789");
var img = Pix.LoadFromMemory(imageStream.GetBuffer());
var page = engine.Process(img);
result = page.GetText();
Console.WriteLine(result);
}
return(String.IsNullOrWhiteSpace(result) ? "Ocr is finished. Return empty" : result);
}
public OcrManager()
{
_tesseract = new TesseractEngine(LangPath, LngStr);
_tesseract.SetVariable("tessedit_char_blacklist",
"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz〈ヽ〉'〝<〟‥“\\ゐ=`”_.丿″\"");
_tesseract.DefaultPageSegMode = PageSegMode.SingleBlock;
}
static Tuple <string, string, string> OCRBitmapV3(Bitmap fullBmp)
{
string textResult = null;
string blocksResult = null;
string exceptionString = null;
return(new Tuple <string, string, string>(textResult, blocksResult, exceptionString));
try
{
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist", "0123456789B");
using (var page = engine.Process(fullBmp))
{
textResult = page.GetText();
blocksResult = "";
using (var iter = page.GetIterator())
{
iter.Begin();
do
{
do
{
do
{
do
{
if (iter.IsAtBeginningOf(PageIteratorLevel.Block))
{
blocksResult += ConsoleWriteLine("<BLOCK>");
}
blocksResult += ConsoleWrite(iter.GetText(PageIteratorLevel.Word));
blocksResult += ConsoleWrite("_");
if (iter.IsAtFinalOf(PageIteratorLevel.TextLine, PageIteratorLevel.Word))
{
blocksResult += ConsoleWriteLine("%");
}
} while (iter.Next(PageIteratorLevel.TextLine, PageIteratorLevel.Word));
if (iter.IsAtFinalOf(PageIteratorLevel.Para, PageIteratorLevel.TextLine))
{
blocksResult += ConsoleWriteLine("§");
}
} while (iter.Next(PageIteratorLevel.Para, PageIteratorLevel.TextLine));
} while (iter.Next(PageIteratorLevel.Block, PageIteratorLevel.Para));
} while (iter.Next(PageIteratorLevel.Block));
}
}
}
}
catch (Exception e)
{
exceptionString = e.ToString();
}
return(new Tuple <string, string, string>(textResult, blocksResult, exceptionString));
}
/// <summary>
/// 获取验证码并自动识别
/// </summary>
/// <returns></returns>
public string GetVerficar(string path)
{
var result = "";
httpItem.URL = $"{urlbase}/verficar.do";
httpItem.ResultType = ResultType.Byte;
httpItem.Cookie = cookie;
httpResult = httpHelper.GetHtml(httpItem);
var image = byteArrayToImage(httpResult.ResultByte);
image.Save(path + "code.bmp");
cookie = httpResult?.Cookie.Replace("; Path=/erp", "").Trim();
using (var engine = new TesseractEngine(path + "tessdata", "eng", EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist", "0123456789");
using (var pix = PixConverter.ToPix((Bitmap)image))
{
using (var page = engine.Process(pix))
{
result = page.GetText();
}
}
}
result = result.Replace("\n", "").Replace(" ", "").Trim();
var timeSpan = (DateTime.Now.AddMilliseconds(expire * 1000) - DateTime.Now);
RedisService.Instance.StringSet($"{RedisPrimaryKey.WebCrawlingCookie}/{uuid}", cookie, timeSpan);
return(result);
}
public override string RecognizeText(IList <Image> images, string lang)
{
string tessdata = Path.Combine(basedir, TESSDATA);
using (TesseractEngine engine = new TesseractEngine(tessdata, lang, EngineMode.Default))
{
engine.SetVariable("tessedit_create_hocr", Hocr ? "1" : "0");
Tesseract.PageSegMode psm = (PageSegMode)Enum.Parse(typeof(PageSegMode), PageSegMode);
StringBuilder strB = new StringBuilder();
int pageNum = 0;
foreach (Image image in images)
{
pageNum++;
using (Pix pix = ConvertBitmapToPix(image))
{
using (Page page = engine.Process(pix, psm))
{
string text = Hocr ? page.GetHOCRText(pageNum - 1) : page.GetText();
if (text == null)
{
return(String.Empty);
}
strB.Append(text);
}
}
}
return(strB.ToString().Replace("\n", Environment.NewLine));
}
}
private static string GetCaptchaText(string captchaFilePath)
{
string captchaText = null;
Pix captcha = null;
try
{
captcha = Pix.LoadFromFile(captchaFilePath);
}
catch (Exception e)
{
Log.Error(e, "Error loading captcha file");
}
if (captcha != null)
{
var grayCaptcha = captcha.ConvertRGBToGray();
var binarizedCaptcha = grayCaptcha.BinarizeSauvolaTiled(10, 0.75f, 1, 2);
var engine = new TesseractEngine(Path.GetFullPath("tessdata"), "eng", EngineMode.Default);
engine.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ1234567890");
var page = engine.Process(binarizedCaptcha, PageSegMode.SparseText);
captchaText = page.GetText();
captchaText = Regex.Replace(captchaText, @"\s+", string.Empty);
}
return(captchaText);
}
public static string OCR(Bitmap img)
{
TesseractEngine ocr = null;
string sResult = "";
try
{
ocr = new TesseractEngine("./tessdata", "eng"); //初始化 (一定要放在tessdata資料夾下)
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); //強迫Char List，較準確
Page page = ocr.Process(img, PageSegMode.SingleLine);
sResult = page.GetText();//result
page.Dispose();
}
catch (Exception ex)
{
//MessageBox.Show(ex.Message);
sResult = "";
}
finally
{
ocr?.Dispose();
}
return(sResult.Replace(" ", ""));
}
// TessNet2 is based on Tesseract v2.04 and has not been updated since September 2009.
// Tesseract 3 .NET wrapper is available here: https://github.com/charlesw/tesseract
private string executeOCR_By_tesseract(Image srcImg)
{
try
{
//var img = Pix.LoadFromFile(imgPath);
//var srcImg = System.Drawing.Image.FromFile(imgPath);
var img = scaleImage(srcImg, 2.3, 2); // Scale up and extend the canvas to get a better result
srcImg.Dispose();
tessEngine.SetVariable("tessedit_char_whitelist", "0123456789:"); // Digits & colons only
//tessEngine.DefaultPageSegMode = PageSegMode.SingleWord; // Without this, the text may not be recognized at all (because of the narrow page margin)
var page = tessEngine.Process(img, PageSegMode.SingleWord); // 如果使用SingleBlock, 识别结果中可能包含空格
var text = page.GetText().Trim();
page.Dispose();
img.Dispose();
Console.WriteLine(text);
return(text);
}
catch (Exception e)
{
Console.WriteLine("Tesseract error: " + e.ToString());
return("Tesseract error: " + e.ToString());
}
}
public Result <char> Process(Mat input)
{
try
{
var buff = new VectorOfByte();
CvInvoke.Imencode(".tiff", input, buff);
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default))
{
engine.DefaultPageSegMode = PageSegMode.SingleChar;
engine.SetVariable("tessedit_char_whitelist", "0123456789ABEKMHOPCTYXDI");
using (var img = Pix.LoadTiffFromMemory(buff.ToArray()))
{
using (var page = engine.Process(img))
{
var text = page.GetText()[0];
_debugLogger.Log(debugLogBuilder => debugLogBuilder.AddMessage("Letter").AddImage(input).AddMessage($"has been recognized as: {text}"));
return(Result.Ok(text));
}
}
}
}
catch (Exception e)
{
return(Result.Fail <char>(e.Message));
}
}
public static string ocrDigitLine(Bitmap source, int x, int y, int width, int height) // 한줄씩 읽어내고 "-" --> "." 으로 치환하고, 공백 제거등, 투약량에 쓰기 좋다.
{
string text;
// var PrescriptionImage = CropedPrescription;
using (var engine = new TesseractEngine(@"C:\Program Files\Tesseract-OCR\tessdata\", "kor", EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist", "0123456789-."); // 숫자와 . - 만 인식하도록 설정
// using (var img = Pix.LoadFromFile(PrescriptionImage))
// {
var roi = new Rect(x, y, width, height); // region of interest 좌표를 생성하고
using (var page = engine.Process(source, roi, PageSegMode.SingleLine)) // psm 옵션 설정
{
text = page.GetText();
text = TextProcess.RemoveWhiteSpace(text);
// text = TextProcess.DotReplace(text);
// text = TextProcess.PointInsert(text);
// Console.WriteLine("인식한 문자: \n{0}\n", text);
// Console.Read();
}
// }
}
return(text);
}
public void CanSetDoubleVariable(string variableName, double variableValue)
{
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default)) {
var variableWasSet = engine.SetVariable(variableName, variableValue);
Assert.That(variableWasSet, Is.True, "Failed to set variable '{0}'.", variableName);
double result;
if(engine.TryGetDoubleVariable(variableName, out result)) {
Assert.That(result, Is.EqualTo(variableValue));
} else {
Assert.Fail("Failed to retrieve value for '{0}'.", variableName);
}
}
}
private static void SetVariablesAccordingToConfig(TesseractEngine engine, ZoneConfiguration barcodeConfig)
{
if (barcodeConfig.TextualDataFilter.FilterType == FilterType.Alpha)
{
engine.SetVariable("tessedit_char_whitelist", "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ-");
}
if (barcodeConfig.TextualDataFilter.FilterType == FilterType.AlphaOnly)
{
engine.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ-");
}
if (barcodeConfig.TextualDataFilter.FilterType == FilterType.Digits)
{
engine.SetVariable("tessedit_char_whitelist", "0123456789");
}
if (barcodeConfig.TextualDataFilter.FilterType == FilterType.Number)
{
engine.SetVariable("tessedit_char_whitelist", "0123456789,.");
}
}
public void CanSetBooleanVariable(bool variableValue)
{
const string VariableName = "classify_enable_learning";
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default)) {
var variableWasSet = engine.SetVariable(VariableName, variableValue);
Assert.That(variableWasSet, Is.True, "Failed to set variable '{0}'.", VariableName);
bool result;
if(engine.TryGetBoolVariable(VariableName, out result)) {
Assert.That(result, Is.EqualTo(variableValue));
} else {
Assert.Fail("Failed to retrieve value for '{0}'.", VariableName);
}
}
}
public void WritesOutThresholdedImageWhenOCRing()
{
var expectedFilePath = Path.Combine(Environment.CurrentDirectory, "tessinput.tif");
if (File.Exists(expectedFilePath)) {
File.Delete(expectedFilePath);
}
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default)) {
engine.SetVariable("tessedit_write_images", true);
using (var img = Pix.LoadFromFile("./phototest.tif")) {
using (var page = engine.Process(img)) {
var region1Text = page.GetText();
Assert.That(File.Exists(expectedFilePath));
}
}
}
}
public void CanSetClassifyBlnNumericModeVariable()
{
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default)) {
engine.SetVariable("classify_bln_numeric_mode", 1);
using (var img = Pix.LoadFromFile("./Data/processing/numbers.png")) {
using (var page = engine.Process(img)) {
var text = page.GetText();
const string expectedText = "1234567890\n\n";
Assert.That(text, Is.EqualTo(expectedText));
}
}
}
}
