static void Main(string[] args)
{
Img();
try
{
string path = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "20.png");
Bitmap image = new Bitmap(path); //识别图像
tessnet2.Tesseract ocr = new tessnet2.Tesseract(); //声明一个OCR类
//ocr.SetVariable("tessedit_char_whitelist", "0123456789+-="); //设置识别变量，当前只能识别数字。
//ocr.SetVariable("tessedit_char_whitelist", "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ");
//ocr.SetVariable("language_model_penalty_non_freq_dict_word", "0");
//ocr.SetVariable("language_model_penalty_non_dict_word ", "0");
//ocr.SetVariable("tessedit_char_blacklist", "xyz");
//ocr.SetVariable("classify_bln_numeric_mode", "1");
string language = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "Language");
ocr.Init(language, "eng", false); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty); //执行识别操作
foreach (tessnet2.Word word in result)
{
Console.WriteLine("{0} : {1}", word.Confidence, word.Text);
}
}
catch (Exception ex)
{
throw;
}
}
/// <summary>
/// Do the actual OCR reading
/// </summary>
/// <remarks>For good results, make sure:
/// The text is large enough
/// The background and foreground of the text are in contrast, and each has its same color
/// Try to give a bitmap with less "noise" as possible
/// </remarks>
/// <returns>the text that was extracted from the image</returns>
public string ProcessImage()
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
if (allowedChars != string.Empty)
{
ocr.SetVariable("tessedit_char_whitelist", allowedChars);
}
ocr.Init(ocrPath + @"\tessdata", "eng", false);
List <tessnet2.Word> result = ocr.DoOCR(imageToRead, Rectangle.Empty);
StringBuilder builder = new StringBuilder();
int lastLine = 1;
foreach (tessnet2.Word word in result)
{
if (word.Confidence < 255.0 * (120.0 - accuracy) / 100.0)
{
if (lastLine < word.LineIndex)
{
builder.Append("\n");
lastLine = word.LineIndex;
}
builder.Append(word.Text + " ");
}
}
return(builder.ToString());
}
/// <summary>
/// 获取验证码
/// </summary>
/// <param name="url"></param>
/// <returns></returns>
public static string GetValidateCode(string url, CookieContainer webCookie, out Bitmap bmpImage)
{
try
{
//获取验证码
Image HeadImage = HttpHelps.GetPicture(url, webCookie);
//处理图片
Bitmap bitmap = new Bitmap(HeadImage);
UnCodebase ud = new UnCodebase(bitmap);
ud.GrayByPixels();
ud.ClearNoise(128, 2);
//识别验证码
tessnet2.Tesseract ocr = new tessnet2.Tesseract(); //声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); //设置识别变量，当前只能识别数字。
ocr.Init(Application.StartupPath + @"\\tmpe", "eng", true); //应用当前语言包。
List <tessnet2.Word> result = ocr.DoOCR(ud.bmpobj, Rectangle.Empty); //执行识别操作
string validateCode = result[0].Text;
//pictureBox1.Image = ud.bmpobj;
bmpImage = ud.bmpobj;
return(validateCode);
}
catch (Exception ex)
{
function.log("验证码获取错误" + ex.Message);
bmpImage = null;
return("");
}
}
public string RecognizeText(IList <Image> images, int index, string lang)
{
using (tessnet2.Tesseract ocr = new tessnet2.Tesseract())
{
ocr.Init(lang, false);
IList <Image> workingImages;
if (index == -1)
{
workingImages = images; // all images
}
else
{
workingImages = new List <Image>();
workingImages.Add(images[index]); // specific image
}
StringBuilder strB = new StringBuilder();
foreach (Bitmap image in workingImages)
{
// If the OcrDone delegate is not null then this'll be the multithreaded version
//ocr.OcrDone = new tessnet2.Tesseract.OcrDoneHandler(Finished);
// For event to work, must use the multithreaded version
//ocr.ProgressEvent += new tessnet2.Tesseract.ProgressHandler(ProgressEvent);
m_event = new ManualResetEvent(false);
List <tessnet2.Word> result = ocr.DoOCR(image, rect);
// Wait here it's finished
//m_event.WaitOne();
if (result == null)
{
return(String.Empty);
}
for (int i = 0; i < tessnet2.Tesseract.LineCount(result); i++)
{
strB.AppendLine(tessnet2.Tesseract.GetLineText(result, i));
}
//int lineIndex = 0;
//foreach (tessnet2.Word word in result)
//{
// if (lineIndex != word.LineIndex)
// {
// strB.AppendLine();
// lineIndex = word.LineIndex;
// }
// strB.Append(new string(' ', word.Blanks)).Append(word.Text);
//}
//strB.AppendLine();
}
return(strB.ToString());
}
}
public List<tessnet2.Word> DoOcrNormal(Bitmap image, string lang)
{
var ocr = new tessnet2.Tesseract();
ocr.Init(_tessdataPath, lang, false);
var result = ocr.DoOCR(image, Rectangle.Empty);
DumpResult(result);
return result;
}
public List<tessnet2.Word> DoOCRNormal(Bitmap image, string lang)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(null, lang, false);
List<tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
DumpResult(result);
return result;
}
public List <tessnet2.Word> DoOCRNormal(Bitmap image, string lang)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(null, lang, false);
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
return(result);
}
public List <tessnet2.Word> DoOCRNormal(Bitmap image, string lang)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(@"C:\Users\mkmak\Documents\Visual Studio 2013\Projects\ConsoleApplication1\tessdata", lang, false);
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
DumpResult(result);
return(result);
}
private static void ProcessImg(Bitmap image)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789");
ocr.Init(@"C:\Users\Emil-PC\Documents\Visual Studio 2015\Projects\kNN\kNN\bin\Debug\tessdata", "eng", false);
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
string resultString = "";
Dictionary <string, List <Rectangle> > CharLocations = new Dictionary <string, List <Rectangle> >();
int charCount = 0;
foreach (tessnet2.Word word in result)
{
//Console.WriteLine(word.ToString());
resultString += word.Text + " ";
foreach (tessnet2.Character character in word.CharList)
{
charCount++;
Rectangle charPosition = FindCharLocation(character.Left, character.Right, character.Top, character.Bottom);
//Console.WriteLine("{0} : {1}", character.Value.ToString(), charPosition.ToString());
List <Rectangle> allCharBounds;
if (!CharLocations.TryGetValue(character.Value.ToString(), out allCharBounds))
{
allCharBounds = new List <Rectangle>();
CharLocations.Add(character.Value.ToString(), allCharBounds);
}
allCharBounds.Add(charPosition);
//G.DrawRectangle(Pens.Blue, charPosition);
}
}
foreach (var charPositions in CharLocations)
{
foreach (var charLocation in charPositions.Value)
{
using (Bitmap croppedImage = ScaleImage(image.Clone(charLocation, image.PixelFormat), scale, scale))
{
//Console.WriteLine("Char: " + charPositions.Key);
FindkNN(k, kNNGroups, croppedImage, charPositions.Key);
}
}
}
foreach (var res in UltimateResult.OrderByDescending(key => key.Value))
{
//Console.WriteLine(res.Key + ": " + Math.Round(res.Value / CharLocations.Count(), 3) * 100); //
Console.WriteLine(res.Key + ": " + Math.Round(res.Value / (k * charCount), 3) * 100);
}
UltimateResult.Clear();
}
/// <summary>
///
/// </summary>
/// <param name="image"></param>
/// <param name="lang"></param>
/// <returns></returns>
public List <tessnet2.Word> DoOcrNormal(Bitmap image, string lang)
{
var ocr = new tessnet2.Tesseract();
ocr.Init(_tessdataPath, lang, false);
var result = ocr.DoOCR(image, Rectangle.Empty);
DumpResult(result);
return(result);
}
public override string ToString()
{
var ocr = new tessnet2.Tesseract();
ocr.Init(null, "eng", false);
ocr.SetVariable("tessedit_char_whitelist", "0123456789,$");
var result = ocr.DoOCR(this.bmp, Rectangle.Empty);
return(result[0].Text.Replace("$", ""));
}
public void DoOcrMultiThred(Bitmap image, string lang)
{
var ocr = new tessnet2.Tesseract();
ocr.Init(_tessdataPath, lang, false);
// If the OcrDone delegate is not null then this'll be the multithreaded version
ocr.OcrDone = Finished;
// For event to work, must use the multithreaded version
ocr.ProgressEvent += OcrProgressEvent;
_mEvent = new ManualResetEvent(false);
ocr.DoOCR(image, Rectangle.Empty);
// Wait here it's finished
_mEvent.WaitOne();
}
public void DoOCRMultiThred(Bitmap image, string lang)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(null, lang, false);
// If the OcrDone delegate is not null then this'll be the multithreaded version
ocr.OcrDone = new tessnet2.Tesseract.OcrDoneHandler(Finished);
// For event to work, must use the multithreaded version
ocr.ProgressEvent += new tessnet2.Tesseract.ProgressHandler(ocr_ProgressEvent);
m_event = new ManualResetEvent(false);
ocr.DoOCR(image, Rectangle.Empty);
// Wait here it's finished
m_event.WaitOne();
}
public void DoOCRMultiThred(Bitmap image, string lang)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(null, lang, false);
// If the OcrDone delegate is not null then this'll be the multithreaded version
ocr.OcrDone = new tessnet2.Tesseract.OcrDoneHandler(Finished);
// For event to work, must use the multithreaded version
ocr.ProgressEvent += new tessnet2.Tesseract.ProgressHandler(ocr_ProgressEvent);
m_event = new ManualResetEvent(false);
ocr.DoOCR(image, Rectangle.Empty);
// Wait here it's finished
m_event.WaitOne();
}
public void Read()
{
////do ocr
foreach (string file in Directory.GetFiles(iDir))
{
if (file.EndsWith(".tif", true, null) || file.EndsWith(".tiff", true, null))// || file.EndsWith(".pdf", true, null))
{
string fileName = file.Substring(file.LastIndexOf("\\"));
fileName = fileName.Substring(0, fileName.LastIndexOf("."));
Bitmap image = new Bitmap(file);
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(null, "ENG", false);
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
image = ocr.GetThresholdedImage(image, Rectangle.Empty); ////
fileName += ".bmp"; ////
image.Save(fileName); ////
for (int i = 0; i < oDir.Length; i++)
{
if (oType[i] == "PDF")
{
fileName += ".pdf";
////create pdf
}
else if (oType[i] == "DOC")
{
fileName += ".doc";
////create doc
}
if (oEmail[i])
{
try
{
SendMail(Properties.Settings.Default.SMTPServer, Properties.Settings.Default.SMTPPort, Properties.Settings.Default.SSL, Properties.Settings.Default.Username, Properties.Settings.Default.Password, Properties.Settings.Default.FromAddress, Properties.Settings.Default.FromName, Properties.Settings.Default.Subject, Properties.Settings.Default.Message, Properties.Settings.Default.HTML, new string[] { oDir[i] }, new string[] { fileName });
File.Delete(fileName);
}
catch (Exception ex)
{
System.Windows.Forms.MessageBox.Show(ex.ToString(), "Error", System.Windows.Forms.MessageBoxButtons.OK, System.Windows.Forms.MessageBoxIcon.Error);
}
}
else
{
File.Move(fileName, oDir[i] + fileName);
}
}
}
}
}
/// <summary>
///
/// </summary>
/// <param name="image"></param>
/// <param name="lang"></param>
public void DoOcrMultiThred(Bitmap image, string lang)
{
var ocr = new tessnet2.Tesseract();
ocr.Init(_tessdataPath, lang, false);
// If the OcrDone delegate is not null then this'll be the multithreaded version
ocr.OcrDone = Finished;
// For event to work, must use the multithreaded version
ocr.ProgressEvent += OcrProgressEvent;
_mEvent = new ManualResetEvent(false);
ocr.DoOCR(image, Rectangle.Empty);
// Wait here it's finished
_mEvent.WaitOne();
}
public void pdf()
{
Bitmap image = new Bitmap("eurotext.tif");
Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); // If digit only
ocr.Init(@"c:\temp", "fra", false); // To use correct tessdata
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
foreach (tessnet2.Word word in result)
{
Console.WriteLine("{0} : {1}", word.Confidence, word.Text);
}
}
public void DoOCRMultiThred(Bitmap image, string lang)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789");
var path = string.Concat(Application.StartupPath, @"\tessdata");
ocr.Init(path, lang, false);
// If the OcrDone delegate is not null then this'll be the multithreaded version
ocr.OcrDone = new tessnet2.Tesseract.OcrDoneHandler(Finished);
// For event to work, must use the multithreaded version
ocr.ProgressEvent += new tessnet2.Tesseract.ProgressHandler(ocr_ProgressEvent);
m_event = new ManualResetEvent(false);
ocr.DoOCR(image, Rectangle.Empty);
// Wait here it's finished
m_event.WaitOne();
}
private void button1_Click(object sender, EventArgs e)
{
string strPath = textBox1.Text;
Bitmap image = new Bitmap(strPath);//识别图像
tessnet2.Tesseract ocr = new tessnet2.Tesseract();//声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); //设置识别变量，当前只能识别数字。
ocr.Init(@"D:\tessdata", "eng", false); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
List<tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);//执行识别操作
string strShow = "";
foreach (tessnet2.Word word in result)
{
strShow = string.Format("{0} : {1}", word.Confidence, word.Text);
MessageBox.Show(strShow);
}
}
public void DoOCRMultiThred(Bitmap image, string lang)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789");
var path = string.Concat(Application.StartupPath, @"\tessdata");
ocr.Init(path, lang, false);
// If the OcrDone delegate is not null then this'll be the multithreaded version
ocr.OcrDone = new tessnet2.Tesseract.OcrDoneHandler(Finished);
// For event to work, must use the multithreaded version
m_event = new ManualResetEvent(false);
ocr.DoOCR(image, Rectangle.Empty);
// Wait here it's finished
m_event.WaitOne();
}
static Tuple <string, int> OCRBitmapV2(Bitmap bSprite)
{
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "0123456789");
ocr.Init(@"", "eng", false);
List <tessnet2.Word> result = ocr.DoOCR(bSprite, Rectangle.Empty);
var resultText = string.Join("", result.Select(i => i.Text));
var resultValue = -101;
if (int.TryParse(resultText, out resultValue))
{
}
return(new Tuple <string, int>(resultText, resultValue));
}
private void processarToolStripMenuItem_Click(object sender, EventArgs e)
{
var image = (Bitmap)currentImage.Clone();
var ocr = new tessnet2.Tesseract();
ocr.Init(@"tessdata", "eng", false);
ocr.SetVariable("tesseract_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVXWYZ-1234567890");
var result = ocr.DoOCR(image, Rectangle.Empty);
StringBuilder sb = new StringBuilder();
foreach (tessnet2.Word word in result)
{
sb.Append(word.Text + " ");
}
MessageBox.Show(String.Format(sb.ToString()));
}
private void button2_Click(object sender, EventArgs e)
{
var image = (Bitmap)pictureBox1.Image;
var ocr = new tessnet2.Tesseract();
ocr.Init(@"tessdata", "eng", true);
ocr.SetVariable("tesseract_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVXWYZ-1234567890");
var result = ocr.DoOCR(image, Rectangle.Empty);
StringBuilder sb = new StringBuilder();
foreach (tessnet2.Word word in result)
{
sb.Append(word.Text + " ");
}
textBox1.Text = sb.ToString();
}
private static void GetValue(HttpHelper httpHelper, string url)
{
var bytes = httpHelper.HttpByteGet(url, string.Empty, false, false, 60 * 1000);
var image = Bitmap(bytes);
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
string language = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "Language");
ocr.Init(language, "eng", false);
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
foreach (tessnet2.Word word in result)
{
Console.WriteLine("{0} : {1}", word.Confidence, word.Text);
}
}
/// <summary>
/// Gets a screenshot of the active window (already optimised for OCR),
/// calculates the thresholded image,
/// runs the OCR,
/// and returns the result as a context entry.
/// </summary>
/// <param name="ce"></param>
/// <returns></returns>
public ContextEntry RunOcr(ContextEntry ce)
{
if (_tEngine == null)
{
return(null);
}
try
{
// run OCR preprocessing
RunOcrPreProcessing(ce.Screenshot);
// processed screenshot
var processedScreenshot = ce.Screenshot.Image;
// threshold
//screenshot = _tEngine.GetThresholdedImage(screenshot, Rectangle.Empty); //TODO: enable or disable?
//Screenshot.SaveImage(screenshot, "thresholded"); //TEMP
_tEngine.OcrDone += OcrFinished;
//var result = _tEngine.DoOCR(screenshot, Rectangle.Empty); // used for single-threading OCR processing
_tEngine.DoOCR(processedScreenshot, Rectangle.Empty);
processedScreenshot.Dispose();
_mEvent = new ManualResetEvent(false);
_mEvent.WaitOne(); // wait here until it's finished
// add ocr'd text to context entry
ce.OcrText = _temporaryOcrText;
ce.Confidence = _tempConfidence;
// reset temp values
_temporaryOcrText = string.Empty;
_tempConfidence = Settings.OcrConfidenceAcceptanceThreshold;
// release sources
ce.Screenshot.Dispose();
ce.Screenshot = null;
}
catch (Exception e)
{
Logger.WriteToLogFile(e);
}
return(ce);
}
private void button1_Click(object sender, EventArgs e)
{
string strPath = textBox1.Text;
Bitmap image = new Bitmap(strPath); //识别图像
tessnet2.Tesseract ocr = new tessnet2.Tesseract(); //声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); //设置识别变量，当前只能识别数字。
ocr.Init(@"D:\tessdata", "eng", false); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty); //执行识别操作
string strShow = "";
foreach (tessnet2.Word word in result)
{
strShow = string.Format("{0} : {1}", word.Confidence, word.Text);
MessageBox.Show(strShow);
}
}
private void processo2()
{
var img = (Bitmap)pictureBox1.Image.Clone();
var showImg = (Bitmap)pictureBox1.Image.Clone();
//img = Grayscale.CommonAlgorithms.Y.Apply(img);
img = new OtsuThreshold().Apply(img);
img = new Erosion().Apply(img);
img = new Invert().Apply(img);
BlobCounter bc = new BlobCounter();
bc.BackgroundThreshold = Color.Black;
bc.ProcessImage(img);
Rectangle rect = new Rectangle(0, 0, showImg.Width, showImg.Height);
BitmapData bmpData = showImg.LockBits(rect, ImageLockMode.ReadWrite, showImg.PixelFormat);
bc.GetObjectsRectangles().ToList().ForEach(i =>
{
Crop filter = new Crop(new Rectangle(i.X, i.Y, 230, 75));
var img2 = (Bitmap)filter.Apply(img);
img2 = new Invert().Apply(img2);
var ocr = new tessnet2.Tesseract();
ocr.SetVariable("tesseract_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVXWYZ-1234567890");
ocr.Init(@"tessdata", "eng", false);
var result = ocr.DoOCR(img2, Rectangle.Empty);
StringBuilder sb = new StringBuilder();
foreach (tessnet2.Word word in result)
{
sb.Append(word.Text + " ");
}
//cliente para servidor
string aux;
aux = sb.ToString();
if (aux.Length >= 6)
{
placas.Add(aux);
}
pictureBox1.Image = img2;
});
showImg.UnlockBits(bmpData);
pictureBox1.Image = img;
}
public string RecognizeText(IList<Image> images, string lang)
{
using (tessnet2.Tesseract ocr = new tessnet2.Tesseract())
{
ocr.Init(null, lang, false);
StringBuilder strB = new StringBuilder();
foreach (Bitmap image in images)
{
// If the OcrDone delegate is not null then this'll be the multithreaded version
//ocr.OcrDone = new tessnet2.Tesseract.OcrDoneHandler(Finished);
// For event to work, must use the multithreaded version
//ocr.ProgressEvent += new tessnet2.Tesseract.ProgressHandler(ProgressEvent);
m_event = new ManualResetEvent(false);
List<tessnet2.Word> result = ocr.DoOCR(image, rect);
// Wait here it's finished
//m_event.WaitOne();
if (result == null) return String.Empty;
for (int i = 0; i < tessnet2.Tesseract.LineCount(result); i++)
{
strB.AppendLine(tessnet2.Tesseract.GetLineText(result, i));
}
//int lineIndex = 0;
//foreach (tessnet2.Word word in result)
//{
// if (lineIndex != word.LineIndex)
// {
// strB.AppendLine();
// lineIndex = word.LineIndex;
// }
// strB.Append(new string(' ', word.Blanks)).Append(word.Text);
//}
//strB.AppendLine();
}
return strB.ToString();
}
}
public async Task <OCRResult> UploadFile()
{
// Verify that this is an HTML Form file upload request
if (!Request.Content.IsMimeMultipartContent("form-data"))
{
throw new HttpResponseException(Request.CreateResponse(HttpStatusCode.UnsupportedMediaType));
}
// Create a stream provider for setting up output streams
MultipartFormDataStreamProvider streamProvider = new MultipartFormDataStreamProvider(ServerUploadFolder);
// Read the MIME multipart asynchronously content using the stream provider we just created.
await Request.Content.ReadAsMultipartAsync(streamProvider);
Bitmap image = new Bitmap(streamProvider.FileData.Select(e => e.LocalFileName).First());
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
//ocr.SetVariable("tessedit_char_whitelist", "0123456789"); // If digit only
try
{
string path = System.Web.HttpContext.Current.Server.MapPath("~/tessdata");
ocr.Init(path, "eng", false); // To use correct tessdata
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
OCRResult res = new OCRResult();
res.FileNames = streamProvider.FileData.Select(entry => entry.LocalFileName);
res.RecognizedTextItems = new List <OCRItem>();
foreach (tessnet2.Word word in result)
{
(res.RecognizedTextItems as List <OCRItem>).Add(
new OCRItem()
{
Text = word.Text,
Confidence = word.Confidence
}
);
}
return(res);
}
catch (Exception e)
{
throw e;
}
}
public async Task<OCRResult> UploadFile()
{
// Verify that this is an HTML Form file upload request
if (!Request.Content.IsMimeMultipartContent("form-data"))
{
throw new HttpResponseException(Request.CreateResponse(HttpStatusCode.UnsupportedMediaType));
}
// Create a stream provider for setting up output streams
MultipartFormDataStreamProvider streamProvider = new MultipartFormDataStreamProvider(ServerUploadFolder);
// Read the MIME multipart asynchronously content using the stream provider we just created.
await Request.Content.ReadAsMultipartAsync(streamProvider);
Bitmap image = new Bitmap(streamProvider.FileData.Select(e => e.LocalFileName).First());
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
//ocr.SetVariable("tessedit_char_whitelist", "0123456789"); // If digit only
try
{
string path = System.Web.HttpContext.Current.Server.MapPath("~/tessdata");
ocr.Init(path, "eng", false); // To use correct tessdata
List<tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
OCRResult res = new OCRResult();
res.FileNames = streamProvider.FileData.Select(entry => entry.LocalFileName);
res.RecognizedTextItems = new List<OCRItem>();
foreach (tessnet2.Word word in result)
{
(res.RecognizedTextItems as List<OCRItem>).Add(
new OCRItem()
{
Text = word.Text,
Confidence = word.Confidence
}
);
}
return res;
}
catch (Exception e)
{
throw e;
}
}
private void button1_Click(object sender, EventArgs e)
{
System.Drawing.Bitmap img = Run();
UnCodebase ud = new UnCodebase(img);
img = ud.GrayByPixels();
ud.ClearNoise(128, 2);
pictureBox1.Image = img;
tessnet2.Tesseract ocr = new tessnet2.Tesseract(); //声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); //设置识别变量，当前只能识别数字。
ocr.Init(Application.StartupPath + @"\\tmpe", "eng", true); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
List <tessnet2.Word> result = ocr.DoOCR(img, Rectangle.Empty); //执行识别操作
string code = result[0].Text;
textBox1.Text = code;
}
public static Point?Buscar(string texto)
{
using (Bitmap bmp = GetCapture())
{
var ocr = new tessnet2.Tesseract();
ocr.Init("tesseract", "spa", false);
var result = ocr.DoOCR(bmp, Rectangle.Empty);
foreach (tessnet2.Word word in result)
{
if (word.Text.Equals(texto))
{
return(new Point(word.Right, word.Bottom));
}
}
}
return(null);
}
private string GetText(Image img, tessnet2.Tesseract ocr, Rectangle cropRect)
{
Bitmap src = new Bitmap(img);
Bitmap target = new Bitmap(cropRect.Width, cropRect.Height);
using (Graphics g = Graphics.FromImage(target))
{
g.DrawImage(src, new Rectangle(0, 0, target.Width, target.Height),
cropRect,
GraphicsUnit.Pixel);
}
// target = make_bw(target);
target = ResizeBitmap(target, target.Width * 10, target.Height * 10);
List <tessnet2.Word> result = ocr.DoOCR(target, Rectangle.Empty);
return(result[0].Text);
}
static void Main(string[] args)
{
try
{
Bitmap image = new Bitmap(@"C:\Users\Nino\Documents\PbpHackthone\dotNet_OCR_server\images\IMAG0253_cropped.jpg");
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); // If digit only
ocr.Init(@"C:\Users\Nino\Desktop\tessdata", "eng", false); // To use correct tessdata
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
foreach (tessnet2.Word word in result)
{
Console.WriteLine("{0} : {1}", word.Confidence, word.Text);
}
}
catch (Exception ex)
{
Console.WriteLine(ex.Message);
}
Console.Read();
}
private void btn_start_Click(object sender, EventArgs e)
{
try
{
/*
var engine = new TesseractEngine(path, "eng", EngineMode.Default);
var page = engine.Process(image);
output = page.GetText();
rich_status.Text += output;
rich_status.Text += recognize(output);
rich_status.Text += "\nagi here" + checkHasDate(output) + getDate(output);
*/
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz.,$-/#&=()'");
ocr.Init(path + @"\tessdata", "eng", false);
List<tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
foreach (tessnet2.Word word in result)
{
count++;
totalConfidence += word.Confidence;
output = string.Join(" ", result.Select(x => x.Text).ToList());
rich_status.Text = output;
using (StreamWriter writetext = new StreamWriter(path + @"\Results.txt"))
{
writetext.WriteLine(output);
}
}
rich_status.Text += recognize(output);
txt_date.Text = getDate(output);
txt_type.Text = getCertificateType(output);
totalConfidence = (totalConfidence / count);
txt_confidence.Text = totalConfidence.ToString();
}
catch (Exception ex)
{
}
}
public static string TryOcr(this Image bitmap)
{
Bitmap temp = new Bitmap(bitmap);
temp = temp.GrayByPixels().ClearNoise(128, 2);
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"); // If digit only
ocr.Init(@"lib\tessdata", "eng", false); // To use correct tessdata
List <tessnet2.Word> result = ocr.DoOCR(temp, Rectangle.Empty);
foreach (tessnet2.Word word in result)
{
SimpleConsole.WriteLine(string.Format("{0} : {1}", word.Confidence, word.Text));
}
if (result.Count > 0)
{
return(result[0].Text);
}
return(string.Empty);
}
public string GetValidateCode(string base64Image)
{
try
{
foreach (var item in base64ImageStart)
{
base64Image = base64Image.Replace(item, "");
}
//过滤特殊字符即可
string dummyData = base64Image.Replace("%0A", "").Replace("%0D", "");
byte[] arr = Convert.FromBase64String(dummyData);
MemoryStream ms = new MemoryStream(arr);
Bitmap bmp = new Bitmap(ms);
ms.Close();
/*Form frm1 = new Form();
* frm1.BackgroundImageLayout = ImageLayout.Zoom;
* frm1.BackgroundImage = bmp;
* frm1.Show();
* UnCodebase codebase = new UnCodebase(bmp);
* Bitmap bitmap = codebase.GrayByPixels(); //灰度处理
* codebase.ClearNoise();//128, 2
* Form frm2 = new Form();
* frm2.BackgroundImageLayout = ImageLayout.Zoom;
* frm2.BackgroundImage = bitmap;
* frm2.Show();*/
tessnet2.Tesseract ocr = new tessnet2.Tesseract();//声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz");
ocr.Init(Application.StartupPath + "\\tessdata", "eng", false);
List <tessnet2.Word> result = ocr.DoOCR(bmp, Rectangle.Empty);//执行识别操作
bmp.Dispose();
string res = result[0].Text;
return(res);
}
catch (Exception ex)
{
return(null);
}
}
public override DecoderItem Decode(byte[] input, VisualDecoderSettings settings)
{
var result = new DecoderItem();
var language = settings.OCRLanguage.ToString();
//use eng if the user dont know the language
if (language.Equals(VisualDecoderSettings.OCRLanguages.unkown.ToString()))
{
language = VisualDecoderSettings.OCRLanguages.eng.ToString();
}
using (var ocr = new tessnet2.Tesseract()){
try
{
ocr.Init(TessractData, language, settings.NumericMode);
List <tessnet2.Word> r1 = ocr.DoOCR(ByteArrayToImage(input), Rectangle.Empty);
//aggregate resultTest
string resultText = "";
int lc = tessnet2.Tesseract.LineCount(r1);
for (int i = 0; i < lc; i++)
{
resultText += tessnet2.Tesseract.GetLineText(r1, i) + "\n";
}
//fill result
result.CodePayload = resultText;
result.CodeType = "none";
result.BitmapWithMarkedCode = input;
ocr.Clear();
}
catch (Exception) // well, the ocr lib sucks... it sometimes trows memory leaks.
// But it is the best opensource lib available.
{
}
}
return(result);
}
private static int[,] RecognizeDigits()
{
int[,] sudoku = new int[9, 9];
for (int i = 0; i < 9; ++i)
{
for (int j = 0; j < 9; ++j)
{
Bitmap image = new Bitmap(outDir + i.ToString() + j.ToString() + ".png");
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(tessData, "eng", false); // To use correct tessdata
List <tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);
foreach (tessnet2.Word word in result)
{
Regex regex = new Regex(@"\d+");
Match match = regex.Match(word.Confidence < CONFIDENCE ? word.Text : "");
sudoku[i, j] = match.Value == "" ? 0 : Int32.Parse(match.Value);
Console.WriteLine("{0} : {1}", word.Confidence, word.Text);
}
}
}
return(sudoku);
}
public string RecognizeText(IList<Image> images, int index, string lang, BackgroundWorker worker, DoWorkEventArgs e)
{
// Abort the operation if the user has canceled.
// Note that a call to CancelAsync may have set
// CancellationPending to true just after the
// last invocation of this method exits, so this
// code will not have the opportunity to set the
// DoWorkEventArgs.Cancel flag to true. This means
// that RunWorkerCompletedEventArgs.Cancelled will
// not be set to true in your RunWorkerCompleted
// event handler. This is a race condition.
this.worker = worker;
if (worker.CancellationPending)
{
e.Cancel = true;
return String.Empty;
}
using (tessnet2.Tesseract ocr = new tessnet2.Tesseract())
{
ocr.Init(lang, false);
IList<Image> workingImages;
if (index == -1)
{
workingImages = images; // all images
}
else
{
workingImages = new List<Image>();
workingImages.Add(images[index]); // specific image
}
StringBuilder strB = new StringBuilder();
foreach (Bitmap image in workingImages)
{
// If the OcrDone delegate is not null then this'll be the multithreaded version
//ocr.OcrDone = new tessnet2.Tesseract.OcrDoneHandler(Finished);
// For event to work, must use the multithreaded version
//ocr.ProgressEvent += new tessnet2.Tesseract.ProgressHandler(ProgressEvent);
m_event = new ManualResetEvent(false);
List<tessnet2.Word> result = ocr.DoOCR(image, rect);
// Wait here it's finished
//m_event.WaitOne();
if (result == null) return String.Empty;
for (int i = 0; i < tessnet2.Tesseract.LineCount(result); i++)
{
strB.AppendLine(tessnet2.Tesseract.GetLineText(result, i));
}
//int lineIndex = 0;
//foreach (tessnet2.Word word in result)
//{
// if (lineIndex != word.LineIndex)
// {
// strB.AppendLine();
// lineIndex = word.LineIndex;
// }
// strB.Append(new string(' ', word.Blanks)).Append(word.Text);
//}
//strB.AppendLine();
}
return strB.ToString();
}
}
private void BtnRun_Click(object sender, EventArgs e)
{
this.webBrowser1.Document.GetElementById("mailNum").InnerText = AppInputParameters.Current.EMSCode ;
HtmlElementCollection ret = webBrowser1.Document.All.GetElementsByName("checkCode");
Image img = GetRegCodePic(this.webBrowser1, "checkCode", null, null);
if (img == null) return;
System.Drawing.Bitmap bitmap = new Bitmap(img);
UnCodebase ud = new UnCodebase(bitmap);
Bitmap processedImg = ud.GrayByPixels();
ud.ClearNoise(128, 2);
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.SetVariable("tessedit_char_whitelist", "0123456789");
ocr.Init(Application.StartupPath + @"\\tmpe", "eng", true);
List<tessnet2.Word> result = ocr.DoOCR(processedImg, Rectangle.Empty);
string code = result[0].Text;
foreach (HtmlElement item in ret)
{
item.InnerText = code;
}
HtmlElement head = webBrowser1.Document.GetElementsByTagName("head")[0];
HtmlElement scriptEl = webBrowser1.Document.CreateElement("script");
IHTMLScriptElement element = (IHTMLScriptElement)scriptEl.DomElement;
element.text = "function SingleFormSubmit() { $(\"*\").hide(); $(\"#singleForm\").submit();}";
head.AppendChild(scriptEl);
webBrowser1.Document.InvokeScript("SingleFormSubmit");
}
private string GetCheckCode(string strPath)
{
string strShow = "";
lock(m_hLockTable)
{
Bitmap image = new Bitmap(strPath);//识别图像
tessnet2.Tesseract ocr = new tessnet2.Tesseract();//声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789"); //设置识别变量，当前只能识别数字。
string strDirect = System.AppDomain.CurrentDomain.BaseDirectory;
strDirect += "tessdata";
ocr.Init(strDirect, "eng", false); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
List<tessnet2.Word> result = ocr.DoOCR(image, Rectangle.Empty);//执行识别操作
foreach (tessnet2.Word word in result)
{
//strShow = string.Format("{0} : {1}", word.Confidence, word.Text);
strShow += word.Text.Trim();
}
image.Dispose();
}
return strShow;
}
public string getCheckStr(string web, CookieContainer cookie)
{
string checkStr = "";
string s = "";
string checkUrl = web + @"/index.php?m=Public&a=verify_code";
HttpWebRequest getCheckWebRequest = (HttpWebRequest)HttpWebRequest.Create(checkUrl);
getCheckWebRequest.Method = "GET";
getCheckWebRequest.Host = "www.nmgqxfw.com";
getCheckWebRequest.KeepAlive = true;
getCheckWebRequest.Accept = @"text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8";
getCheckWebRequest.UserAgent = @"Mozilla/5.0 (Windows NT 6.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/32.0.1700.102 Safari/537.36";
getCheckWebRequest.Referer = web + @"/system.php?m=Public&a=index";
getCheckWebRequest.Headers.Add(@"Accept-Encoding", @"gzip,deflate,sdch");
getCheckWebRequest.Headers.Add(@"Accept-Language", @"zh-CN,zh;q=0.8,en;q=0.6");
getCheckWebRequest.CookieContainer = cookie;
Bitmap img, imgsource;
using (HttpWebResponse webResponse = (HttpWebResponse)getCheckWebRequest.GetResponse())
{
using (Stream respStream = webResponse.GetResponseStream())
{
byte[] buffer = new byte[100000];
int read, total = 0;
while ((read = respStream.Read(buffer, total, 1000)) != 0)
{
total += read;
}
imgsource = (Bitmap)Bitmap.FromStream(new MemoryStream(buffer, 0, total));
img = imgsource.Clone(new System.Drawing.Rectangle(0, 0, imgsource.Width, imgsource.Height), PixelFormat.Format24bppRgb);
respStream.Close();
}
}
try
{
img.Save(@"E:\迅雷下载\zdyzmsb\BMP\origin.bmp");
img = BlackAndWhite(img, 0.8);
//img = new Bitmap(img, img.Width * 3, img.Height * 3);
//Color originC;
//int whit = 255, black = 0;
//int nearVal = 0;
//int MaxNearPoints = 3;
//int width = img.Width;
//int height = img.Height;
//byte[] p = new byte[9];
//for (int x = 0; x < width; x++)
//{
// for (int y = 0; y < height; y++)
// {
// originC = img.GetPixel(x, y);
// if (!(255 == originC.R && 255 == originC.G && 255 == originC.B))
// {
// img.SetPixel(x, y, Color.FromArgb(originC.A, black, black, black));
// nearVal = 0;
// if (0 == x || img.Width - 1 == x || 0 == y || img.Height - 1 == y)
// {
// img.SetPixel(x, y, Color.FromArgb(originC.A, whit, whit, whit));
// }
// else
// {
// if (black == img.GetPixel(x - 1, y - 1).R) nearVal++;
// if (black == img.GetPixel(x, y - 1).R) nearVal++;
// if (black == img.GetPixel(x + 1, y - 1).R) nearVal++;
// if (black == img.GetPixel(x - 1, y).R) nearVal++;
// if (black == img.GetPixel(x + 1, y).R) nearVal++;
// if (black == img.GetPixel(x - 1, y + 1).R) nearVal++;
// if (black == img.GetPixel(x, y + 1).R) nearVal++;
// if (black == img.GetPixel(x + 1, y + 1).R) nearVal++;
// }
// if (nearVal < MaxNearPoints)
// img.SetPixel(x, y, Color.FromArgb(originC.A, whit, whit, whit));
// }
// }
//}
////img.Save("black.png");
//Bitmap img1 = img.Clone(new Rectangle(0, 0, 16, img.Height), PixelFormat.Format24bppRgb);
//img1.Save("1.png");
//Bitmap img2 = img.Clone(new Rectangle(17, 0, 17, img.Height), PixelFormat.Format24bppRgb);
//img2.Save("2.png");
//Bitmap img3 = img.Clone(new Rectangle(35, 0, 17, img.Height), PixelFormat.Format24bppRgb);
//img3.Save("3.png");
//Bitmap img4 = img.Clone(new Rectangle(53, 0, 17, img.Height), PixelFormat.Format24bppRgb);
//img4.Save("4.png");
tessnet2.Tesseract ocr = new tessnet2.Tesseract();//声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"); //设置识别变量，当前只能识别数字。
ocr.Init(@"E:\Projects\CSharp\Test\Test\Lan", "eng", true); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
//List<tessnet2.Word> result = ocr.DoOCR(img, Rectangle.Empty);//执行识别操作
Monitor.Enter(this);//因为是多线程，所以用到了Monitor。
System.Collections.Generic.List<tessnet2.Word> result = ocr.DoOCR(img, System.Drawing.Rectangle.Empty);//执行识别操作
foreach (tessnet2.Word word in result) //遍历识别结果。
s = s + word.Text;
Monitor.Exit(this);
if (s.Length > 2)
s = s.Substring(2, s.Length - 2);
checkStr = result[0].Text;
}
catch
{
}
return checkStr;
}
public Dictionary<string, string> UseTessnet2(string imgFile, Dictionary<string, Rectangle> pDicOcrRect)
{
this.ErrMsg = string.Empty;
string defLang = "eus";
Dictionary<string, string> dicResult = new Dictionary<string, string>();
if (pDicOcrRect == null)
{
pDicOcrRect = new Dictionary<string, Rectangle>();
pDicOcrRect.Add(this.emptyRCKey, Rectangle.Empty);
}
tessnet2.Tesseract ocr = new tessnet2.Tesseract();
ocr.Init(this.TESSDATA, defLang, false);
using (Bitmap m_image = new Bitmap(imgFile))
{
foreach (var recItem in pDicOcrRect)
{
Rectangle rcTemp = recItem.Value;
if (rcTemp != Rectangle.Empty)
{
int dx = Convert.ToInt16(m_image.HorizontalResolution / 100);
int dy = Convert.ToInt16(m_image.VerticalResolution / 100);
rcTemp = new Rectangle(recItem.Value.X * dx, recItem.Value.Y * dy,
recItem.Value.Width * dx, recItem.Value.Height * dy);
}
List<tessnet2.Word> m_words = ocr.DoOCR(m_image, rcTemp);
StringBuilder sbWords = new StringBuilder();
foreach (var item in m_words)
{
sbWords.AppendLine(item.Text);
}
string text = sbWords.ToString();
if (!string.IsNullOrEmpty(text))
{
//// correct common errors caused by OCR
//text = TextUtilities.CorrectOCRErrors(text);
//// correct letter cases
//text = TextUtilities.CorrectLetterCases(text);
text = text.Replace("\n", Environment.NewLine);
}
else
{
text = string.Empty;
}
dicResult.Add(recItem.Key, text);
}
}
return dicResult;
}
private void button1_Click(object sender, EventArgs e)
{
Bitmap bitmap = (Bitmap) Bitmap.FromFile("OKK.png");
UnCodebase ud = new UnCodebase(bitmap);
bitmap = ud.GrayByPixels();
ud.ClearNoise(128, 2);
pictureBox1.Image = bitmap;
tessnet2.Tesseract ocr = new tessnet2.Tesseract();//声明一个OCR类
ocr.SetVariable("tessedit_char_whitelist", "0123456789ABCDEFGHIGKLMNOPQRSTUVWXYZ"); //设置识别变量，当前只能识别数字。
string path = Application.StartupPath + @"\tmpe\";
ocr.Init(path, "eng", false); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
List<tessnet2.Word> result = ocr.DoOCR(bitmap, Rectangle.Empty);//执行识别操作
string code = result[0].Text;
textBox1.Text = code;
return;
//System.Drawing.Bitmap img = Run();
//UnCodebase ud = new UnCodebase(img);
//img = ud.GrayByPixels();
//ud.ClearNoise(128, 2);
//pictureBox1.Image = img;
//tessnet2.Tesseract ocr = new tessnet2.Tesseract();//声明一个OCR类
//ocr.SetVariable("tessedit_char_whitelist", "0123456789"); //设置识别变量，当前只能识别数字。
//ocr.Init(Application.StartupPath + @"\\tmpe", "eng", true); //应用当前语言包。注，Tessnet2是支持多国语的。语言包下载链接：http://code.google.com/p/tesseract-ocr/downloads/list
//List<tessnet2.Word> result = ocr.DoOCR(img, Rectangle.Empty);//执行识别操作
//string code = result[0].Text;
//textBox1.Text = code;
}