ImageExtract product guide · Updated September 20, 2026
Decide what you need to recover
If you need a photograph without the surrounding text, try embedded-image extraction. If you need an exact view of the complete page, use page rendering in a PDF application. The two operations produce different outputs.
Why a page can contain several images
A PDF may store a photograph, its transparency mask, a small logo, and decorative elements independently. Extraction can return several fragments instead of one page-sized picture. Vector charts and text are not automatically embedded images.
Scanned documents behave differently
A scanned page is often stored as one raster image. Extracting that image may recover the whole scan, including its text. It does not recognize the text or isolate every photograph within the scan.
Inspect before using the output
Upload a readable PDF, run extraction, then compare the preview with the document. Check dimensions, color, transparency, and format. Document processing can convert images; do not assume a downloaded preview is byte-identical to the embedded source.
When a preview looks wrong
Complex color spaces and masks can produce unexpected results. Keep the source file and compare with a trusted PDF viewer. A known issue with unusually dark previews in some documents remains under investigation; do not use affected output as a faithful reproduction.