[
https://issues.apache.org/jira/browse/PDFBOX-3854?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16075188#comment-16075188
]
Tilman Hausherr edited comment on PDFBOX-3854 at 7/5/17 6:55 PM:
-----------------------------------------------------------------
It is really just that method. I am testing copying the image the hard way
(without ColorConvertOp) and that works fast too, but I need to rerun all tests
again.
{code}
public BufferedImage toRGBImage(WritableRaster raster) throws IOException
{
init();
ColorModel colorModel = new ComponentColorModel(awtColorSpace,
false, false, Transparency.OPAQUE,
raster.getDataBuffer().getDataType());
BufferedImage image = new BufferedImage(colorModel, raster, false,
null);
BufferedImage dest = new BufferedImage(image.getWidth(),
image.getHeight(), BufferedImage.TYPE_INT_RGB);
int width = image.getWidth();
int height = image.getHeight();
for (int x = 0; x < width; ++x)
{
for (int y = 0; y < height; ++y)
{
dest.setRGB(x, y, image.getRGB(x, y));
}
}
return dest;
}
{code}
What I am wondering about: in my 1000 PDF test I didn't notice any improvement,
maybe this is because they are very different. Is your project something where
all PDF files are always with RGB colorspaces?
was (Author: tilman):
It is really just that method. I am testing copying the image the hard way
(without ColorConvertOp) and that works fast too, but I need to rerun all tests
again.
What I am wondering about: in my 1000 PDF test I didn't notice any improvement,
maybe this is because they are very different. Is your project something where
all PDF files are always with RGB colorspaces?
> PDDeviceRGB.toRGBImage does not return a RGB-Typed-Image
> --------------------------------------------------------
>
> Key: PDFBOX-3854
> URL: https://issues.apache.org/jira/browse/PDFBOX-3854
> Project: PDFBox
> Issue Type: Bug
> Affects Versions: 2.0.6
> Reporter: Yannick Haser
> Attachments: rgbImageExample0.pdf, rgbImageSecondExample.pdf,
> Starter2.java, Starter.java
>
>
> While converting a PDPage to a BufferedImage (using the PDFRenderer)
> Images-Resources located in the PDPage gets loaded in a WriteableRaster and a
> given Colorspace.
> These get converted to a BufferedImage (with PDDeviceRGB.toRGBImage()). But
> the Type of the returing BufferedImage is with the given Raster and
> Colorspace always from Type "BufferedImage.TYPE_CUSTOM" instead of the
> expected "BufferedImage.TYPE_INT_RGB" for example.
> Consequences:
> Image-Processing-Operations with/on BufferedImages with TYPE_CUSTOM are not
> optimized in Java (slower) and partialy runs through "Synchronized" Blocks.
> ==> Converting different Documents in different Threads are blocking each
> other.
> "Quick"-Fix for seeing a Performance-Improvement in a Multithreading
> Environment:
> {code:title=PDDeviceRGB.java|borderStyle=solid}
> @Override
> public BufferedImage toRGBImage(WritableRaster raster) throws IOException {
> init();
> ColorModel colorModel = new ComponentColorModel(awtColorSpace, false,
> false, Transparency.OPAQUE,
> raster.getDataBuffer().getDataType());
> BufferedImage image = new BufferedImage(colorModel, raster, false,
> null);
> BufferedImage dest = new BufferedImage(image.getWidth(),
> image.getHeight(), BufferedImage.TYPE_INT_RGB);
> ColorConvertOp op = new
> ColorConvertOp(dest.getColorModel().getColorSpace(), null);
> return op.filter(image, dest);
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
