Am 23.02.2018 um 08:31 schrieb Itai:

I'm trying to use PDFBox to show a preview of some PDFs containing very
large images (typically tens of thousands of pixels a side).
It seems PDFRenderer uses PDImage#toImage, which creates a BufferedImage at
the full resolution, regardless of the DPI or scale requested by
This causes an issue, as the resulting BufferedImage is often too large to
fit in memory (>8GB), causing an OOM error.

Is there a way to get a scaled BufferedImage at a lower quality, without
first having to render/produce a full-resolution BufferedImage?


If not - where would I start if I needed to add such a feature?

You could grab the source code and look at it... that isn't so easy. You could look at SampledImageReader and debug from there. The problem is that this class wouldn't know that a "smaller" image is needed.


To unsubscribe, e-mail:
For additional commands, e-mail:

Reply via email to