Extract Images

What does "Extract images" save?

The pictures stored inside the PDF — photos, scans, logos — one file per picture. A plain JPEG is saved exactly as it is stored in the PDF (so there is no quality loss), with any camera or location data removed (a JPEG whose metadata cannot be removed safely is saved as a PNG of its pixels instead, never with the location data left in). Every other kind — including a JPEG that has transparency or CMYK colour, where passing the bytes through would change how it looks — is saved as a PNG of the decoded pixels. A picture used on several pages is listed once.

Why is an image missing, or marked "cannot be extracted"?

Several things are not listed as pictures. Vector art (charts, logos made of shapes) is not an image — use PDF to Image to turn whole pages into pictures. JPEG 2000, JBIG2 and CCITT fax images — common in scanned documents — need a decoder HerePDF’s browser engine does not include here, so they are listed by name with no download rather than saved wrong. Pictures that no page draws are not listed, and neither are pictures drawn only inside a tiling pattern, a Type 3 font glyph or a comment or form-field appearance, nor small “inline” images written directly into a page’s drawing instructions; the few shapes called stencil masks (a colour poured through a pattern) are counted but not saved.

Will the saved picture match what I see on the page?

It is the stored picture, not a screenshot of the page: it is not cropped, rotated or scaled the way the page shows it, and any text or marks drawn over it are not included. A picture with transparency is saved as a PNG with its transparency. If you want the page as it looks, use PDF to Image.

Does it keep the location data in my photos?

No. A JPEG is saved with its camera and GPS metadata removed; its pixels are not touched. A colour profile that the PDF attaches to the picture from outside the JPEG file is not added to the saved file, so a wide-gamut photo may look slightly different in another viewer. The PDF itself is not changed, and nothing is uploaded.

Why does it refuse some PDFs?

If the author of a PDF has turned off copying in its permissions, HerePDF honours that and will not extract its images — the same rule as Extract Text and PDF to Image. A PDF that needs a password to open asks for it first.

How many images can I save at once?

"Download all as ZIP" saves every picture it can extract, one after another. A document with hundreds of large scans needs memory to hold the ZIP before it is saved; if your device struggles, download pictures one at a time.

What happens with a very large picture?

A JPEG, or a picture stored as plain RGB or grey pixels, is normally saved at any size. A picture of another kind has to be drawn through your browser first, and phones in particular limit how large that can be (about 16.7 million pixels). A picture above the limit is named as too large to decode and has no download, rather than being saved wrong.