public void getText2(string inputFile)
{
// Open a PDF file.
String inputFilePath = inputFile;
PDFDocument doc = new PDFDocument(inputFilePath);
// The folder that contains '.traineddata' files.
OCRHandler.SetTrainResourcePath(@"D:\Alice\DLL\Source\");
// Set output file path.
Stream[] streams = new MemoryStream[doc.GetPageCount()];
for (int i = 0; i < doc.GetPageCount(); i++)
{
BasePage page = doc.GetPage(i);
streams[i] = new MemoryStream();
//the default resolution is 96, if you set larger, it will be helpful to recognize the text, but it can't be too large.
Bitmap bmp = page.ConvertToImage(96);//192,288....
OCRPage ocrPage = OCRHandler.Import(bmp);
ocrPage.Recognize();
ocrPage.SaveTo(MIMEType.DOCX, streams[i]);
streams[i].Seek(0, SeekOrigin.Begin);
}
DOCXDocument.CombineDocument(streams, @"C:\output.docx");
}
public bool doOCR(string fid)
{
Response.ContentType = "text/plain";
Response.Clear();
try
{
int pageIndex = Convert.ToInt32(Request.Params["pageindex"].ToString());
string language = Request.Params["lang"];
string filename = Request.Params["filename"];
string projectName = HttpContext.Current.Request.PhysicalApplicationPath.Replace("\\", "/");
int x = (int)Convert.ToDouble(Request.Params["x"].ToString());
int y = (int)Convert.ToDouble(Request.Params["y"].ToString());
int width = (int)Convert.ToDouble(Request.Params["w"].ToString());
int height = (int)Convert.ToDouble(Request.Params["h"].ToString());
string resourcepath = projectName + "/OCRSource/";
OCRHandler.SetTrainResourcePath(resourcepath);
List <Language> langs = new List <Language>();
if (language.Equals("English"))
{
langs.Add(Language.Eng);
}
else if (language.Equals("German"))
{
langs.Add(Language.Deu);
}
else if (language.Equals("French"))
{
langs.Add(Language.Fra);
}
else if (language.Equals("Dutch"))
{
langs.Add(Language.Nld);
}
else if (language.Equals("Italian"))
{
langs.Add(Language.Ita);
}
else if (language.Equals("Portuguese"))
{
langs.Add(Language.Por);
}
else if (language.Equals("Spanish"))
{
langs.Add(Language.Spa);
}
else if (language.Equals("Arabic"))
{
langs.Add(Language.Ara);
}
Object fileObject = RasterEdge.WDP.Handler.FileinfoHandler.LoadFile(fid, filename, "");
OCRHandler.Settings.LanguagesEnabled = langs;
OCRPage ocrPage = null;
Bitmap image = null;
if (fileObject is BaseDocument)
{
BaseDocument doc = (BaseDocument)fileObject;
BasePage page = doc.GetPage(pageIndex);
image = page.ConvertToImage(1.0f);
ocrPage = OCRHandler.Import(image);
}
else if (fileObject is REImage)
{
image = ((REImage)fileObject).Bitmap;
}
if (image != null)
{
Bitmap newImg = new Bitmap(width, height);
Graphics g = Graphics.FromImage(newImg);
g.DrawImage(image, 0, 0, new Rectangle(x, y, newImg.Width, newImg.Height), GraphicsUnit.Pixel);
ocrPage = OCRHandler.Import(newImg);
ocrPage.Recognize();
g.Dispose();
newImg.Save(@"d:\1.png");
newImg.Dispose();
image.Dispose();
string text = ocrPage.GetText();
text = text.Replace("\r\n", "<br/>");
text = text.Replace("\"", "\\" + "\"");
Response.Write(text);
}
else
{
Response.Write("None");
}
}
catch (System.Exception ex)
{
Response.Write("None");
}
Response.End();
return(true);
}
