To delete a photo from a PDF while keeping the page structure, use Remove Images from PDF to create a local copy in which eligible embedded JPG image objects are replaced by white pixels. The workflow is document-wide rather than a click-and-delete selection: the tool checks every indirect image object in one PDF, substitutes each eligible object with a valid 1 by 1 white JPEG, and leaves the page's drawing command and resource reference in place. This suppresses an eligible visible photograph or scan without rasterizing the whole page. It is designed to retain surrounding text, links, page geometry, and non-image vector content. Choose one local PDF no larger than 25 MB, select Remove eligible images to run the compatibility check, then inspect every page before downloading the new copy. The check is all-or-nothing. If even one embedded image uses an unsupported form, the tool stops without saving an output instead of quietly leaving that image in a partly edited document. This local method is appropriate when a white area is acceptable; it is not a certified redaction workflow for legal, security, or privacy-sensitive removal.
What the white replacement actually changes
Deleting a photo from a PDF can describe several different jobs: erasing a picture, removing the page that carries it, or redacting information. This workflow addresses embedded-image removal while keeping the page. It does not ask you to draw a box around one picture. Once the entire document passes, every eligible image object is replaced, so it is unsuitable when only one of several supported pictures should disappear.
The result is a separate local derivative of the source PDF. Existing page instructions remain in place, but the image object they reference now contains a tiny white JPEG. The page is designed to keep its geometry and does not become one large raster image. Ordinary text and links remain where the source PDF supports them, while non-image vector content is retained. This approach works well for simple placed photographs or scans when the image area is allowed to appear white.
Which images qualify for replacement
The compatibility check examines all indirect image XObjects. For the document to pass, every one must be an unmasked, 8-bit DCTDecode stream using DeviceRGB or DeviceGray, with no DecodeParms. DCTDecode identifies JPEG image data in the supported PDF form. A visually ordinary JPG is not enough on its own; its internal PDF representation must also meet every rule.
| What the check finds | Tool response | What it means |
|---|---|---|
| Unmasked, 8-bit DCTDecode JPEG using DeviceRGB or DeviceGray, with no DecodeParms | Eligible for replacement | The existing image stream is replaced by a 1 by 1 white JPEG. |
| Any one image uses a mask, DecodeParms, another filter, an indexed palette, CMYK, JPX, JBIG2, LZW, or another unsupported form | The entire job stops before saving | No partial PDF is offered as a completed removal. |
| An inline image | Removal is not claimed by this version | Do not assume an inline image will be replaced. |
This strict rule prevents a download from implying that every photo was handled when at least one complex image was left untouched. If the tool reports an unsupported image, there is no output to download. This version does not provide a way to force the document through the check.
Remove the photo in three steps
- Choose one local PDF. Open the tool and select a file from your device. The supported input is one PDF up to 25 MB, and processing takes place locally without a document upload.
- Select Remove eligible images. This starts the all-or-nothing compatibility check. The tool must be able to handle every indirect image object under its safety rules before it can prepare an output file.
- Inspect the new PDF, then download it. Review the pages and confirm that the white replacements are appropriate for your document. Download the new copy only after that inspection.
Keep the original file available for comparison. If any embedded image is unsupported, the tool saves nothing, so there is no partially edited PDF to download. For a successful job, the output is a new derivative rather than a reason to overwrite the only copy of the source.
Why the resource reference remains valid
A PDF page often contains a drawing command that points to an image object by a resource name. Simply deleting that object from the resource dictionary can leave the command referring to something that no longer exists, potentially producing a damaged PDF. Remove Images from PDF deliberately avoids that approach. It retains the original indirect object reference and assigns a replacement PDFRawStream containing the white JPEG, so the page command and resource reference remain valid. The pdf-lib PDFRawStream implementation shows the stream object type used by this process.
After saving the derivative, the browser opens it again with the PDF editor library and then with the site's PDF.js viewer. The verification stage checks that the page count remains the same and that the expected image objects still exist as valid replacements. The Mozilla PDF.js API documents the library used by the PDF viewer. These are structural safeguards, not a promise that every feature of every PDF has been independently validated.
Review every output page
A structurally valid PDF can still need a careful visual review before you rely on it. Compare the new file with the source and check the following:
- Confirm that the page count and intended page geometry remain unchanged.
- Review every location where an eligible photograph or scan appeared. A white replacement can be covered by text or vector content drawn later on the page.
- Check that ordinary text, links, and non-image vector material still look and behave as expected where the source supported them.
- Keep the source document and treat the result as a separate copy rather than an irreversible replacement.
Processing is local: no document upload or server-side edit is involved. Local processing does not, however, turn the result into certified redaction. Retained document metadata or inaccessible original history may still matter in a sensitive workflow, and the white replacement is not a pixel-perfect redaction method. Use a dedicated professional workflow when removal is required for legal, security, or privacy purposes.
Encrypted, malformed, signed, or unusual PDFs may also be rejected. The tool does not bypass a password, preserve a digital signature, or repair corruption. Those limits are important when deciding whether image replacement is an appropriate method for a particular document.
Match the task to the right PDF operation
Remove Images from PDF fits the narrow goal of suppressing supported embedded JPG objects while retaining their pages. If the unwanted item is an entire page, Delete PDF Pages is the more appropriate operation. If the goal is to keep a copy rather than remove the photo, use an image-extraction workflow instead.
| Requested outcome | Most suitable operation | Expected result |
|---|---|---|
| Remove eligible embedded JPG photos or scans while retaining the page | Remove Images from PDF | Eligible image streams become white, but the entire PDF must pass the compatibility check. |
| Remove a full page | Delete PDF Pages | |
| Extract Images from PDF | ||
| Remove sensitive material for legal, security, or privacy purposes | Dedicated redaction workflow | The process is handled separately from ordinary white-image replacement. |
For the exact photo-removal task, choose the white-replacement tool only when a white area is acceptable and the source file meets every compatibility rule. Then inspect every output page before sharing or distributing the new PDF.