Recover a picture without capturing the whole page
A PDF page may combine text, vector drawings and embedded raster images. This tool looks for supported raster image objects and exports their decoded pixels as individual PNG files. It is useful for recovering a photograph from a report or reusing a bitmap diagram when you no longer have the separate image file.
Extraction does not recreate the full page layout. A caption drawn as PDF text, a vector border and a page background are separate objects and may not appear in the extracted PNG. Use the page-to-image converter when you want an image that includes everything visible on a page.
Match the output to what you need
| Your situation | Recommended approach |
|---|---|
| You want a photograph or bitmap used inside a PDF. | Extract the images and compare the thumbnails with the source. The results identify the page where each extracted object was encountered and show its pixel dimensions. |
| You want a complete slide, chart with labels or scanned page as displayed. | Render the page with PDF to Image. An individual embedded object may exclude labels, clipping, rotations or other visual elements added by the PDF page. |
| You need the exact original JPEG or camera file. | Ask for the source asset. This tool creates new PNGs from decoded pixels; it does not return the original compressed byte stream, file format, filename or camera metadata. |
Find the sample's embedded raster image
The downloadable five-page sample contains a raster image alongside ordinary PDF text. It lets you see the difference between recovering an image object and taking a screenshot of its page.
- Choose device-document-lab.pdf, then select Extract embedded images. There is no output quality or page-range setting in this tool.
- Wait for Embedded images ready. Read the image count and the PDF page count separately: these describe different things.
- Select the image thumbnail to download its PNG. Note the source page and dimensions shown below it.
- Open the PNG beside the corresponding PDF page. Compare the picture itself, and notice which page text or surrounding layout was not part of that raster object.
Expected result. The sample should produce a downloadable image while the summary reports five PDF pages. A five-page PDF does not necessarily produce five images: a page can contain no raster objects, one image or several.
Check the downloaded result
- Open each required PNG after download. Compare its subject, crop, transparency and colours with the source page; the thumbnail alone may hide small differences.
- Check pixel dimensions before placing an image in a new document. Enlarging a small embedded picture does not create additional detail.
- Keep the source-page number with any image you reuse, and verify that you have permission to use it. Downloading an embedded object does not establish ownership or a reuse licence.
Troubleshooting
- No standalone embedded raster images were found.
The visible artwork may be vector shapes or text rather than a supported raster image. Very small or oversized image objects can also be skipped by the extractor.
Try this: Use PDF to Image for a complete page rendering. A visually complex page can still contain no image objects that this tool can export.
- The PNG is missing a caption, looks rotated differently or shows more than the page crop.
The page can transform or clip an image and draw text over it. The extracted object's pixels do not include the surrounding page composition.
Try this: Render the complete page if those elements must stay together. Keep an extracted image only when its independent appearance suits your task.
- Extraction stops at an image-count or output-size limit.
Decoded PNGs can use more space than compressed images inside the source PDF. The tool limits the number of images and total output for the current device.
Try this: Extract a smaller group of PDF pages first, then process that file. The selection area shows the current image-count limit; the operation does not offer partial downloads after exceeding it.
Limits and compatibility
- Results are decoded PNGs
- Only supported raster representations are exported. Masks, unusual encodings and page effects can prevent a faithful standalone result. Repeated uses of the same image object may be returned once rather than once per visible occurrence.
- File access and browser resources
- The PDF and exported pixels are processed locally. Password-protected files must be unlocked with permission before extraction. The browser must support canvas image encoding and have enough memory for the selected PDF and resulting images.
Your files and form values are processed on this device. Downloaded copies remain wherever you save them. See how to check local processing and our privacy policy for site data and optional services.
Common questions
- Why did my scanned page become one large image?
- A scan can be stored as a single page-sized raster image. Extraction recovers that object; it does not identify and separate individual photographs within the scan.
- Can I download everything in a ZIP?
- This tool currently provides individual PNG downloads through the result thumbnails. Save each image you need before selecting Start over or leaving the page.
Continue your document workflow
Maintained by Shakti Prasad Ram. Found a mismatch? Report the tool, browser and steps without sending a sensitive document. Back to the tool