[PDFBOX-198] Tiff image problems

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Duplicate
• Fix Version/s: 1.8.5, 2.0.0
• Component/s: Writing
• Assignee: Unassigned
• Reporter: Anonymous
• Created: 15/Sep/06 08:36
• Updated: 02/May/14 06:12
• Resolved: 22/Feb/14 12:50

Description

[imported from SourceForge]
http://sourceforge.net/tracker/index.php?group_id=78314&atid=552832&aid=1559157
Originally submitted by wolf12 on 2006-09-15 01:36.

Hi everyone!!!

PDFBox doesn't write tiff images propertly.
In tiff documentation there are many fields that are
not suported by this library.
1. DPI (XResolution, YResolution, ResoulutionUnit)
XResolution and YResolution IFDFields are hardcoded
(300 dpi value for both), ResolutionUnit is not even
taken into account. User is forced to write those
values by hisself.
2. FillOrder
It is a logical order of bits within a byte (ref.
TIFF6.pdf document page 32). In PDFBox this entry is
allways 1 (because of lack it in library this field has
default value 1 - available are 1 and 2). The efect of
this is that the data is not properly recognized by pdf
viewers - error shown:
"No sufficient data for the image".

Here is code for creating pdf page with tiff image:

1. >>>>>>>>>>> create COSDictionary <<<<<<<<<<<

IFDEntry columns =
ifd.directoryEntries.get(IFDName.ImageWidth);
IFDEntry rows =
ifd.directoryEntries.get(IFDName.ImageLength);
IFDEntry t4compression =
ifd.directoryEntries.get(IFDName.Compression);
IFDEntry blackIs1 =
ifd.directoryEntries.get(IFDName.PhotometricInterpretation);
IFDEntry t42d =
ifd.directoryEntries.get(IFDName.T4Options);

int k = -1000;
if (t4compression != null)
{
if (t4compression.getValue() == 4)

{ k = -1; }

if (t4compression.getValue() == 3)

{ k = 0; }

}
if (t42d != null && t42d.getValue() == 1)

{ k = 50; }

if (k == -1000)

{ throw new IOException("First image in tiff is not CCITT T4 or T6 compressed"); }

boolean bi1 = (blackIs1.getValue() == 1) ? true : false;

cosDictParms.setInt(COLUMNS, columns.getValue());
cosDictParms.setInt(ROWS, rows.getValue());
if (bi1)

{ cosDictParms.setBoolean("BlackIs1", true); }

cosDictParms.setInt("K", k);
cosDictParms.setFloat("XRes", this.getXRes());
cosDictParms.setFloat("YRes", this.getYRes());
this.checkCOSDictionary = true;

2. >>>>>>>>>>> create PDStream <<<<<<<<<<<

PDStream stream = new PDStream(doc);
COSStream dic = stream.getStream();
COSDictionary decodeParms =
tiffImage.getCosDictionaryParams();
OutputStream os = null;
try

{ os = dic.createFilteredStream(); os.write(tiffImage.getStripsByte()); }

finally
{
try
{
if (os != null)

{ os.close(); }

}
catch (IOException e)

{ // ignore error }

}
dic.setItem(COSName.FILTER, COSName.CCITTFAX_DECODE);
dic.setItem(COSName.SUBTYPE, COSName.IMAGE);
dic.setItem(COSName.TYPE, COSName.getPDFName("XObject"));
dic.setItem("DecodeParms", decodeParms);

3. >>>>>>>>>>> create PDXObjectImage <<<<<<<<<<<

COSDictionary decodeParms =
tiffImage.getCosDictionaryParams();
PDCcitt ccitt = new PDCcitt(stream);
ccitt.setBitsPerComponent(1);
ccitt.setColorSpace(new PDDeviceGray());
ccitt.setWidth(decodeParms.getInt(TiffImage.COLUMNS));
ccitt.setHeight(decodeParms.getInt(TiffImage.ROWS));

4. >>>>>>>>>> create pdf page with image <<<<<<<<<

PDPageContentStream contentStream = new
PDPageContentStream(document, page);
PDFPageImageFormat pdfFormat =
PDFPageImageFormat.newPDFPageImageFormat(0, 0,
PDFPageImageFormat.FIT_TO_IMAGE, imageObject);
page.setMediaBox(new
PDRectangle(tiffImage.getOryginalWidth(),
tiffImage.getOryginalHeight()));
contentStream.drawImage(imageObject,
pdfFormat.getX(), pdfFormat.getY(),
(int)tiffImage.getOryginalWidth(),/pdfFormat.getWidth(),/

(int)tiffImage.getOryginalHeight()/pdfFormat.getHeight()/);
bos.write(tiffImage.toByteArray());

>>>>>>>>>>>>> end source code <<<<<<<<<

I cannot attach any example of pdf file because of
security.

I have my own tiff wraper which reads also multipage
tiff image.

Issue Links

• duplicates: PDFBOX-1734 ImageIoUtil.WriteImage doesn't work with tiff images