Any ideas on how to remove images from PDF file?

[tatagi]

I want to keep the pdf file in same layouts and alignments including texts, but without images. the pdf file is not scanned document, it's basically text with images attached.

I know I can manually edit/remove images in online/desktop editors but it takes too long time for hundreds of my files in the queue.

It's discouraging to know PDF isn't as easy to control elements as EPUB is.

Are there easier ways to removes only images using calibre or any other software out there?


converting to TXT isn't an option because it destorys the whole structure of the document.

[Foozle]

Doing a search, I found one program that may work. You can check it out here: VeryPDF PDF to EPUB Converter. It uses a command line to convert and ignore images. There's also Adobe Acrobat itself (or other PDF editors) that can be used to remove images, but that has to be done one file at a time. Otherwise you may be stuck doing it manually as there's no other automagic way (that I've ever personally seen) of stripping images besides just converting / saving to pure text.

[Quoth]

PDF is an end-use format rather than editable so in extreme cases scissors are needed.

Really this sounds like too much work and will only work for some PDFs. I'm curious as to why you want to remove the images?

[DaltonST]

Since Adobe designed PDFs to faithfully and accurately print a document, it has always been a terrible ebook format.

Having said that, it is of course ubiquitous in academic publication databases due to that characteristic, although modern users nowadays don't routinely physically print them. They simply view them, which should be identical to "print preview" by design.

If you cannot actually delete the images (because: see above), perhaps an acceptable alternative would be to use available utilities to reduce the quality of the images, and change them to grayscale.

That would drastically reduce the file size of your PDFs, which is often the motivation to remove the images. It is possible that changing the image attributes might disrupt the "flow" of the PDF document that you prize. Only trial-and-error testing of different combinations of criteria will tell you if that is so.


DaltonST

[Quoth]

Even 4bit or 1 bit greyscale is possible. The 4 bit isn't too bad (14 greys, black, white), but 1 bit only suits line art and equations. Imagemagick or k2pdfopt or other tools.

[tatagi]

THANK YOU ALL FOR THE HEAD-UPS!

Originally Posted by Foozle

Doing a search, I found one program that may work. You can check it out here: VeryPDF PDF to EPUB Converter. It uses a command line to convert and ignore images. There's also Adobe Acrobat itself (or other PDF editors) that can be used to remove images, but that has to be done one file at a time. Otherwise you may be stuck doing it manually as there's no other automagic way (that I've ever personally seen) of stripping images besides just converting / saving to pure text.

Thx. I tried verypdf but unfortunately it just closes its command prompt window without doing anything. PDF is , as others said, the worst ebook format to convert from. I can confirm this after the attempt to convert pdf to epub using Calibre (source document was column separated magazine document and it was all mixed up as one)

Originally Posted by DaltonST

Since Adobe designed PDFs to faithfully and accurately print a document, it has always been a terrible ebook format.

Having said that, it is of course ubiquitous in academic publication databases due to that characteristic, although modern users nowadays don't routinely physically print them. They simply view them, which should be identical to "print preview" by design.

If you cannot actually delete the images (because: see above), perhaps an acceptable alternative would be to use available utilities to reduce the quality of the images, and change them to grayscale.

That would drastically reduce the file size of your PDFs, which is often the motivation to remove the images. It is possible that changing the image attributes might disrupt the "flow" of the PDF document that you prize. Only trial-and-error testing of different combinations of criteria will tell you if that is so.

DaltonST

Thx. you clearly got my point of asking yes it was the file size that I am concerned with, especially for the pdf that contains one hel of a lot of images that I don't need to view. I am usually text-oriented person

It is a nice idea to utterly reduce the quality to the last drop, instead of simply removing them. But Isn't pdf basically unreflowable anyway? If the image in the page is gone, the image is just replaced by empty space and does not affect the page layouts, imho

Originally Posted by Quoth

Even 4bit or 1 bit greyscale is possible. The 4 bit isn't too bad (14 greys, black, white), but 1 bit only suits line art and equations. Imagemagick or k2pdfopt or other tools.

Thx. I will try them out. k2pdfopt seems to serve different purpose (reflow pdf text to fit in smaller screen size by cropping the original pages in half or more) but there're options that fiddle with output image quality.

[Quoth]

Originally Posted by tatagi

Thx. I will try them out. k2pdfopt seems to serve different purpose (reflow pdf text to fit in smaller screen size by cropping the original pages in half or more) but there're options that fiddle with output image quality.

Imagemagic and K2pdfopt can crop. K2pdfopt can do stuff with a text layer. But both can leave the page size as is and change from colour to monochrome and reduce to 4 bits or 1bit (4bits is the minimum I would use).

Most PDFs can't be reflowed anyway.

[ownedbycats]

Originally Posted by DaltonST

Since Adobe designed PDFs to faithfully and accurately print a document, it has always been a terrible ebook format.

There's exceptions to this of course-- some books that rely heavily on layout tend to look better as a fixed-layout pdf than a flow-layout epub. (There's epub3 but Calibre doesn't handle those.)

[Quoth]

Originally Posted by ownedbycats

There's exceptions to this of course-- some books that rely heavily on layout tend to look better as a fixed-layout pdf than a flow-layout epub. (There's epub3 but Calibre doesn't handle those.)

Then they are kinds of books that need large screens and are electronic documents, not ebooks.

I agree there are kinds of books that either:
Useless on a small mono screen (Big books with photos).
Need Print replica on large screen (Many science & maths textbooks).
Or Magazines and newspapers are often unsuitable for small screens.

But PDFs for novels, even with illustrations, are crazy. Trying to perfectly mimic a print edition fiction or poetry or play is also crazy, better to simplify.

[ownedbycats]

I have several books (mixed paper and Kindle-unpacked to PDF) from DK/Dorling Kindersley and they tend to be layout-heavy:



They can be read on eInk, but it's kind of awkward and no colour. I prefer my big monitor.

[BetterRed]

Originally Posted by ownedbycats

There's exceptions to this of course-- some books that rely heavily on layout tend to look better as a fixed-layout pdf than a flow-layout epub. (There's epub3 but Calibre doesn't handle those.)

I have hundreds of EPUB3 books in my calibre libraries, amongst which are many fixed layout books. So I'm unsure what you mean by Calibre doesn't handle them.

I can view them with calibre, albeit with loss of colour in some cases, if that's a problem I can use another viewer via the Open (View) With feature. I can edit them with calibre and/or Sigil, and convert to PDF providing I'm prepared to do any necessary tweaks with a PDF editor.

BR

[HAL.Unke]

With PDF-Xchange Editor (free version) you can copy or remove images from pdfs.
Tools>content editing tools>Edit content>images
Then mark and copy or delete an image.

I use this feature to copy not only the recipes from food magazines. Often are the images without text because they are an extra layer.

[ownedbycats]

Originally Posted by BetterRed

I have hundreds of EPUB3 books in my calibre libraries, amongst which are many fixed layout books. So I'm unsure what you mean by Calibre doesn't handle them.

I can view them with calibre, albeit with loss of colour in some cases, if that's a problem I can use another viewer via the Open (View) With feature. I can edit them with calibre and/or Sigil, and convert to PDF providing I'm prepared to do any necessary tweaks with a PDF editor.

BR

Need a third-party viewer to properly view them, mainly.

How did you convert to PDF? Mine just tend to stall out or produce files that barely work.

[DNSB]

Originally Posted by BetterRed

I can view them with calibre, albeit with loss of colour in some cases, if that's a problem I can use another viewer via the Open (View) With feature. I can edit them with calibre and/or Sigil, and convert to PDF providing I'm prepared to do any necessary tweaks with a PDF editor.

Loss of colour, I could live with. Totally screwed up layouts? That a different kettle of fish. See the attached images for a sample from a FLO epub3 in calibre viewer vs. Sigil's preview vs. Thorium.

[BetterRed]

The few I've converted with calibre, mainly for 'fun', are short papers and so-called roadmaps from a local ¬think-tank. They offer DOCX and EPUB3, but to get a PDF you must jump hurdles. I think they must enjoy whinging subscribers.

I have an admin-tag that acts as a reminder (via an icon) that a book is best read in another viewer, shortcut for AZARDI is Shift+Z.

My point was, that for me at least calibre handles EPUB3 fine, your comment implied 'not at all'.

BR