[PDFBOX-1223] Strange color issues with convertToImage method

• Type: New Feature
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.6.0
• Fix Version/s: 2.0.0
• Component/s: Rendering
• Labels: colorspaces convertToImage overprintcontrol rendering transparency transparent
• Assignee: Andreas Lehmkühler
• Reporter: Jimi Hullegård
• Created: 06/Feb/12 16:42
• Updated: 17/Mar/16 19:06
• Resolved: 17/Dec/14 11:33

Description

While testing PDFBox I noticed a few strange color issues when converting the first page of an pdf to an image using the method convertToImage on a PDPage. An example pdf and resulting png is attached for each case.

• Case 1: Most of the colors seem to get inverted.

• Case 2: The green color of the area in the top right is paler the the original, and the text under the logo at the bottom right is quite "pixely"

• Case 3: The blue area under the text is missing

• Case 4: The red area should be semi-transparent (so the rose stem can be seen behind it), not solid.

I compared all resulting images with how the pdf looks in Foxit Reader 5.1 on Windows 7.

This is the code I used:

---------
PDFParser parser = new PDFParser(new FileInputStream(pdfInputFile));
parser.parse();
PDDocument document = parser.getPDDocument();
List<PDPage> pages = (List<PDPage>)document.getDocumentCatalog().getAllPages();

if (pages.size() > 0) {
PDPage page = pages.get(0);
BufferedImage image = page.convertToImage(BufferedImage.TYPE_INT_RGB, 300);
ImageIO.write(image, "png", pngOutputFile);
}
---------

If there is some change I can do in my code or some configuration to improve some of these issues, then I would love to hear about it. As a comparison, Ghost4j (with Ghostscript underneath) can handle these cases better. Even the text under the logo in case 2 was less "pixely", with the same resolution.

Issue Links

• depends upon: PDFBOX-1169 Images extracted from PDF are loosing color (are shown in blackcolor)