/// <summary>Preprocess given image if it is needed.</summary>
/// <param name="inputImage">
/// original input image
/// <see cref="System.IO.FileInfo"/>
/// </param>
/// <param name="pageNumber">number of page to be OCRed</param>
/// <returns>
/// path to output image as
/// <see cref="System.String"/>
/// </returns>
private String PreprocessImage(FileInfo inputImage, int pageNumber)
{
String tmpFileName = TesseractOcrUtil.GetTempFilePath(Guid.NewGuid().ToString(), GetExtension(inputImage));
String path = inputImage.FullName;
try {
if (GetTesseract4OcrEngineProperties().IsPreprocessingImages())
{
Pix pix = ImagePreprocessingUtil.PreprocessImage(inputImage, pageNumber);
TesseractOcrUtil.SavePixToTempPngFile(tmpFileName, pix);
if (!File.Exists(System.IO.Path.Combine(tmpFileName)))
{
System.Drawing.Bitmap img = TesseractOcrUtil.ConvertPixToImage(pix);
if (img != null)
{
TesseractOcrUtil.SaveImageToTempPngFile(tmpFileName, img);
}
}
}
if (!GetTesseract4OcrEngineProperties().IsPreprocessingImages() || !File.Exists(System.IO.Path.Combine(tmpFileName
)))
{
TesseractOcrUtil.CreateTempFileCopy(path, tmpFileName);
}
if (File.Exists(System.IO.Path.Combine(tmpFileName)))
{
path = tmpFileName;
}
}
catch (System.IO.IOException e) {
LogManager.GetLogger(GetType()).Error(MessageFormatUtil.Format(Tesseract4LogMessageConstant.CANNOT_READ_INPUT_IMAGE
, e.Message));
}
return(path);
}
public virtual void TestReadingInvalidImagePath()
{
NUnit.Framework.Assert.That(() => {
String path = TEST_IMAGES_DIRECTORY + "numbers_02";
FileInfo imgFile = new FileInfo(path);
ImagePreprocessingUtil.PreprocessImage(imgFile, 1);
}
, NUnit.Framework.Throws.InstanceOf <Tesseract4OcrException>())
;
}
/// <summary>
/// Gets OCR result from provided single page image and preprocesses it if
/// it is needed.
/// </summary>
/// <param name="inputImage">
/// input image
/// <see cref="System.IO.FileInfo"/>
/// </param>
/// <param name="outputFormat">
/// selected
/// <see cref="OutputFormat"/>
/// for tesseract
/// </param>
/// <param name="pageNumber">number of page to be OCRed</param>
/// <returns>result as string that will be written to a temporary file later</returns>
private String GetOcrResultForSinglePage(FileInfo inputImage, OutputFormat outputFormat, int pageNumber)
{
String result = null;
try {
// preprocess if required
if (GetTesseract4OcrEngineProperties().IsPreprocessingImages())
{
// preprocess and try to ocr
result = new TesseractOcrUtil().GetOcrResultAsString(GetTesseractInstance(), ImagePreprocessingUtil.PreprocessImage
(inputImage, pageNumber), outputFormat);
}
if (result == null)
{
System.Drawing.Bitmap bufferedImage = ImagePreprocessingUtil.ReadImage(inputImage);
if (bufferedImage != null)
{
try {
result = new TesseractOcrUtil().GetOcrResultAsString(GetTesseractInstance(), bufferedImage, outputFormat);
}
catch (Exception e) {
// NOSONAR
LogManager.GetLogger(GetType()).Info(MessageFormatUtil.Format(Tesseract4LogMessageConstant.CANNOT_PROCESS_IMAGE
, e.Message));
}
}
if (result == null)
{
// perform ocr using original input image
result = new TesseractOcrUtil().GetOcrResultAsString(GetTesseractInstance(), inputImage, outputFormat);
}
}
}
catch (Exception e) {
// NOSONAR
LogManager.GetLogger(GetType()).Error(MessageFormatUtil.Format(Tesseract4LogMessageConstant.TESSERACT_FAILED
, e.Message));
throw new Tesseract4OcrException(Tesseract4OcrException.TESSERACT_FAILED);
}
return(result);
}
