I need to find the (raster) images in a pdf file and resize them (to change their resolution).
My code is based on the example PrintImageLocations. For the part that works, I extend PDFStreamEngine and do special processing for the "Do" operator:
I get the original PDImageXObject using the first parameter of the operator and the resources.
then I create a BufferedImage from that and work on it to change the number of pixels.
then I create a new PDImageXObject from the BufferedImage via LosslessFactory
finally I put the new object in the page resources using the original object's name
I try to do something similar for inline images and I arrive at the point where I have a BufferedImage, but I don't know how to use it to replace the original inline image.
It would also be ok to replace the inline image with a XObject, but again, I don't know how to substitute the two...
Below is my code; the interesting part is the function "processOperator".
// WIP!
// find raster images inside a pdf
// if their resolution is more than 900dpi
// then resize them
// reducing the resolution to 200dpi
// NB bug: fails on pdf files with more than one page
// ...DEBUG ScratchFileBuffer:516 - ScratchFileBuffer not closed!
// also fails on pdf with included pdf
// (e.g. latex \includegraphics{x.pdf})
// # to compile:
// apt install libpdfbox2-java
// export CLASSPATH=.:/usr/share/java/pdfbox2.jar:/usr/share/java/commons-logging.jar
// javac Resampleimages.java
// # to run:
// java Resampleimages x.pdf
// see
// https://pdfbox.apache.org/2.0/examples.html
// https://pdfbox.apache.org/docs/2.0.11/javadocs/
import org.apache.pdfbox.cos.COSBase;
import org.apache.pdfbox.cos.COSName;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.graphics.PDXObject;
import org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject;
import org.apache.pdfbox.pdmodel.graphics.image.PDImage;
import org.apache.pdfbox.pdmodel.graphics.image.PDImageXObject;
import org.apache.pdfbox.pdmodel.graphics.image.PDInlineImage;
import org.apache.pdfbox.util.Matrix;
import org.apache.pdfbox.contentstream.operator.DrawObject;
import org.apache.pdfbox.contentstream.operator.Operator;
import org.apache.pdfbox.contentstream.PDFStreamEngine;
import org.apache.pdfbox.pdmodel.PDResources;
import org.apache.pdfbox.pdmodel.graphics.image.LosslessFactory;
import java.io.File;
import java.io.IOException;
import java.util.List;
import java.awt.image.BufferedImage;
import java.awt.Graphics2D;
import java.awt.RenderingHints;
import java.awt.Color;
import org.apache.pdfbox.contentstream.operator.state.Concatenate;
import org.apache.pdfbox.contentstream.operator.state.Restore;
import org.apache.pdfbox.contentstream.operator.state.Save;
import org.apache.pdfbox.contentstream.operator.state.SetGraphicsStateParameters;
import org.apache.pdfbox.contentstream.operator.state.SetMatrix;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
public class ReplaceBigImages extends PDFStreamEngine
{
private static Log log = LogFactory.getLog(ReplaceBigImages.class);
public ReplaceBigImages() throws IOException
{
addOperator(new Concatenate());
addOperator(new DrawObject());
addOperator(new SetGraphicsStateParameters());
addOperator(new Save());
addOperator(new Restore());
addOperator(new SetMatrix());
}
public static PDDocument document;
public static void main( String[] args ) throws IOException
{
if( args.length != 1 )
{
usage();
}
else
{
try
{
document = PDDocument.load(new File(args[0]));
ReplaceBigImages printer = new ReplaceBigImages();
int pageNum = 0;
for( PDPage page : document.getPages() )
{
pageNum++;
log.info( "Processing page: " + pageNum );
printer.processPage(page);
}
}
finally {
if( document != null )
{
document.save(args[0].replace(".pdf", "_test.pdf"));
document.close();
}
}
}
}
protected void processOperator( Operator operator, List<COSBase> operands) throws IOException
{
String operation = operator.getName();
// log.debug(String.format("Operator %s", operation));
if( "Do".equals(operation) ) {
log.debug("### Found Do operator");
COSName objectName = (COSName) operands.get( 0 );
PDXObject xobject = getResources().getXObject( objectName );
// log.debug(String.format("%s isa %s", objectName, xobject.getClass().getSimpleName()));
if( xobject instanceof PDImageXObject)
{
log.debug(String.format("Looking at %s (%s)", objectName.getName(), xobject));
PDImageXObject image = (PDImageXObject)xobject;
BufferedImage scaledImage = changeImageResolution(image);
if (scaledImage != null) {
log.debug(String.format("Replacing with %s", scaledImage));
PDImageXObject replacement_img = LosslessFactory.createFromImage(document, scaledImage);
PDPage currentPage = getCurrentPage();
PDResources resources = currentPage.getResources();
resources.put(objectName, replacement_img);
}
}else if(xobject instanceof PDFormXObject)
{
PDFormXObject form = (PDFormXObject)xobject;
showForm(form);
}
} else if ("BI".equals(operation)) {
PDPage currentPage = getCurrentPage();
log.debug("### Found BI operator");
PDResources resources = currentPage.getResources();
PDInlineImage image = new PDInlineImage(operator.getImageParameters(),
operator.getImageData(),
resources);
BufferedImage scaledImage = changeImageResolution(image);
if (scaledImage != null) {
log.debug(String.format("Replacing with %s", scaledImage));
PDImageXObject replacement_img = LosslessFactory.createFromImage(document, scaledImage);
// ARGH!!! How do I replace the inline image???
resources.add(replacement_img, "pippo");
// operator.setImageParameters(scaledImage???)
// operator.setImageData(scaledImage???)
}
} else {
super.processOperator( operator, operands);
}
}
protected BufferedImage changeImageResolution( PDImage image)
throws IOException
{
int imageWidth = image.getWidth();
int imageHeight = image.getHeight();
Matrix ctmNew = getGraphicsState().getCurrentTransformationMatrix();
float imageXScale = Math.abs(ctmNew.getScalingFactorX());
float imageYScale = Math.abs(ctmNew.getScalingFactorY());
float resolution = imageWidth / ( imageXScale / 72 );
String stencil = "";
if (image.isStencil()) {
stencil = " (stencil)";
}
// TODO: take into consideration the size at which this file is included by TeX
log.debug("size: ("+imageWidth+","+imageHeight+")@("+imageXScale+","+imageYScale+") resolution = "+resolution+stencil);
// if ( resolution > 899f ) {
if ( resolution > 200f ) {
// what do the following two lines mean???
BufferedImage bImage = new BufferedImage(imageWidth,
imageHeight,
BufferedImage.TYPE_INT_ARGB);
if (image.isStencil()) {
log.warn("Is stencil; painting black.");
bImage = image.getStencilImage(Color.black);
} else {
bImage = image.getImage();
}
int desiredResolution = 200;
float xFactor = (imageXScale / 72) * desiredResolution / imageWidth;
float yFactor = (imageYScale / 72) * desiredResolution / imageHeight;
log.info("Scaling x to "+xFactor);
int dWidth = (int) (xFactor * imageWidth);
int dHeight = (int) (yFactor * imageHeight);
// the image type is from
// https://docs.oracle.com/javase/6/docs/api/constant-values.html#java.awt.image.
log.debug(String.format("Destination: %d x %d [%s]",
dWidth,
dHeight,
bImage.getType()));
BufferedImage scaledImage = new BufferedImage(dWidth,
dHeight,
bImage.getType());
Graphics2D graphics2D = scaledImage.createGraphics();
graphics2D.setRenderingHint(RenderingHints.KEY_INTERPOLATION,
RenderingHints.VALUE_INTERPOLATION_BILINEAR);
graphics2D.setRenderingHint(RenderingHints.KEY_RENDERING,
RenderingHints.VALUE_RENDER_QUALITY);
graphics2D.setRenderingHint(RenderingHints.KEY_ANTIALIASING,
RenderingHints.VALUE_ANTIALIAS_ON);
graphics2D.drawImage(bImage, 0, 0, dWidth, dHeight, null);
graphics2D.dispose();
// see https://pdfbox.apache.org/docs/2.0.11/javadocs/org/apache/pdfbox/pdmodel/graphics/image/PDImageXObject.html#createFromByteArray-org.apache.pdfbox.pdmodel.PDDocument-byte:A-java.lang.String-
return scaledImage;
}
return null;
}
private static void usage()
{
System.err.println( "Usage: java " + ReplaceBigImages.class.getName() + " <input-pdf>" );
}
}
