PDFs can contain images embedded directly in the file structure — photographs, charts, logos, diagrams, illustrations. Extracting them gives you those images as individual files, at their original resolution, without taking screenshots or cropping.
Embedded images vs page renders
There are two ways to get images from a PDF. Extracting embedded images pulls out the image objects that exist in the PDF's binary structure — these are the original images that were placed into the document when it was created, at their original resolution and format.
Rendering the page to an image (PDF to JPG or PDF to PNG) converts the entire page — text, layout, and images together — into a raster image. This is useful when you want a snapshot of the full page, not the individual images.
For recovering specific photos or graphics from a document (a product image from a catalog, a chart from a report, a logo from a business card), image extraction gives you the original file. Page rendering gives you a screenshot.
What gets extracted
The tool extracts all XObject images in the PDF — every image that was embedded as a distinct object. Background images, texture fills, and images that were rasterized into the page content stream (rather than placed as embedded objects) may not be extractable as individual files.
Each extracted image is saved in its original format (JPEG, PNG, or other formats as stored in the PDF). The output is delivered as a ZIP when there is more than one image.
Resolution and quality
Embedded images are extracted at their stored resolution — the quality of each image is exactly what was placed into the PDF. A high-resolution photo embedded at 300 DPI is extracted at 300 DPI. A low-resolution thumbnail embedded at 72 DPI is extracted at 72 DPI. The extraction process does not upscale or re-compress.