private void IterateFullPage(ResultIterator iter, ref List <TextLine> _textLines)
{
int left, top, right, bottom;
StringBuilder ss = new StringBuilder(string.Empty);
PageIteratorLevel level = PageIteratorLevel.RIL_TEXTLINE;
string t;
do
{
TextLine l = new TextLine();
t = iter.GetUTF8Text(level);
ss.Append(t);
iter.BoundingBox(level, out left, out top, out right, out bottom);
l.Bounds = new Rectangle(left, top, right - left, bottom - top);
l.Text = t ?? string.Empty;
level = PageIteratorLevel.RIL_WORD;
l.Words = new List <Word>();
do
{
Word w = new Word();
iter.BoundingBox(level, out left, out top, out right, out bottom);
w.Text = iter.GetUTF8Text(level);
w.Confidence = iter.Confidence(level);
w.Bounds = new Rectangle(left, top, right - left, bottom - top);
l.Words.Add(w);
if (iter.IsAtFinalElement(PageIteratorLevel.RIL_TEXTLINE, PageIteratorLevel.RIL_WORD))
{
break;
}
} while (iter.Next(level));
level = PageIteratorLevel.RIL_TEXTLINE;
ss.Append(System.Environment.NewLine);
_textLines.Add(l);
} while (iter.Next(level));
text = ss.ToString();
}
//public static void clearFaceFlag() { faceFlag = 0; }
public string RecognizeText(int id)
{
DBService dbs = new DBService();
Document doc = dbs.FindDocumentById(id);
Image <Gray, Byte> img = scale(doc);
//var image = PixConverter.ToPix(img.ToBitmap()))
Tesseract.Native.DllImports.TesseractDirectory = System.Web.HttpContext.Current.Server.MapPath("~/Tesseract/bin/Debug/DLLS/");
TessBaseAPI tessBaseAPI = new TessBaseAPI();
System.Diagnostics.Debug.WriteLine("The current version is {0}", tessBaseAPI.GetVersion());
string dataPath = System.Web.HttpContext.Current.Server.MapPath("~/tessdata/");
string language = "eng";
string inputFile = doc.Path;
OcrEngineMode oem = OcrEngineMode.DEFAULT;
//OcrEngineMode oem = OcrEngineMode.DEFAULT;
PageSegmentationMode psm = PageSegmentationMode.AUTO_OSD;
// Initialize tesseract-ocr
if (!tessBaseAPI.Init(dataPath, language, oem))
{
throw new Exception("Could not initialize tesseract.");
}
// Set the Page Segmentation mode
tessBaseAPI.SetPageSegMode(psm);
// Set the input image
Pix pix = tessBaseAPI.SetImage(inputFile);
// Recognize image
tessBaseAPI.Recognize();
ResultIterator resultIterator = tessBaseAPI.GetIterator();
// extract text from result iterator
StringBuilder stringBuilder = new StringBuilder();
int top, bottom, left, right, i = 0;
List <OCRText> forJson = new List <OCRText>();
PageIteratorLevel pageIteratorLevel = PageIteratorLevel.RIL_TEXTLINE;
do
{
string textContent = resultIterator.GetUTF8Text(pageIteratorLevel);
resultIterator.BoundingBox(pageIteratorLevel, out left, out top, out right, out bottom);
string coordsString = "" + left + "," + top + "," + right + "," + bottom;
forJson.Add(new OCRText()
{
Coords = coordsString, Text = textContent
});
} while (resultIterator.Next(pageIteratorLevel));
tessBaseAPI.Dispose();
pix.Dispose();
var textForReturn = JsonConvert.SerializeObject(forJson);
dbs.UpdateDocument(textForReturn, id);
if (HasFace(img) == true)
{
FaceFlag = 1;
}
else
{
FaceFlag = 0;
}
return(textForReturn);
}