Inspect copyright metadata and watermark images with Transloadit

Joseph Grabski

Content Lead · Rochester, United Kingdom · Show bio

Image metadata can help organize a rights-review workflow, but it cannot tell you whether you have permission to publish an image. Missing copyright fields do not make a photograph free to use. For example, the U.S. Copyright Office⁠ explains that copyright protection does not depend on registration. Applicable rights and exceptions depend on the jurisdiction and circumstances.

This is a technical workflow, not a legal determination. Keep evidence of ownership, licenses and permissions separately. Transloadit can route files by metadata and add attribution; it does not provide an automatic copyright-clearance service.

Checking image metadata

The /file/filter Robot can select files whose extracted copyright field is nonempty. This example places those files in a review result; it does not approve or publish them:

{
  "steps": {
    ":original": { "robot": "/upload/handle" },
    "copyright_review": {
      "robot": "/file/filter",
      "use": ":original",
      "accepts": [["${file.meta.copyright}", "!=", ""]],
      "error_on_decline": false,
      "result": true
    }
  }
}

In array filter conditions, missing metadata is compared with an empty string, as documented in the filter reference. A file selected here has a metadata value to inspect, not a verified owner. A file omitted here has no usable value in that field, not permission to publish. Fields vary by format and can be removed or forged.

Review the result metadata in your application and keep uploads private while permissions are unresolved. Do not add an export Step that also reads :original, since that would bypass a filter intended to control publication.

Reverse image search

Reverse image search can help a reviewer find visually similar images and possible sources. A match does not establish ownership, and no match does not establish permission. Review the original source and license evidence yourself.

Use a provider that actually accepts an image or image URL for visual matching. A text-based image search endpoint is not automatically reverse image search, and a search license filter is not a legal verdict on an uploaded file. Transloadit does not provide this search operation directly.

DMCA tooling

Provide an appropriate review and reporting process for your service. Keep the original evidence, review decisions, publication status and relevant permissions auditable. A metadata filter cannot replace a jurisdiction-appropriate notice-handling process. Ask qualified counsel about legal obligations rather than inferring them from this processing example.

How do I watermark my images?

For images you own or are authorized to modify, /image/resize can add a visible watermark. A watermark is a notice or branding element, not proof of ownership or a guarantee against copying.

This Template adds a sample logo and returns only the watermarked result. Replace the logo with artwork you have permission to use:

{
  "steps": {
    ":original": { "robot": "/upload/handle" },
    "watermarked": {
      "robot": "/image/resize",
      "use": ":original",
      "result": true,
      "format": "jpg",
      "watermark_url": "https://demos.transloadit.com/inputs/transloadit-padded.png",
      "watermark_size": "25%",
      "watermark_position": "bottom-right",
      "imagemagick_stack": "v3"
    }
  }
}

Inspect the watermark's size, contrast and placement on representative images. Add a storage export from watermarked only when you are ready to publish; do not include the unwatermarked original in that public export.

Adding copyright metadata to an image

You can also write an attribution notice into a supported file's metadata. The following Assembly Instructions resize and visibly label the image, then write the metadata after that transformation. This avoids losing newly written metadata during a later resize:

{
  "steps": {
    ":original": { "robot": "/upload/handle" },
    "watermarked": {
      "robot": "/image/resize",
      "use": ":original",
      "format": "jpg",
      "width": 1280,
      "height": 720,
      "resize_strategy": "fit",
      "text": [{
        "text": "Copyright ${fields.rights_holder}",
        "size": 24,
        "font": "Ubuntu",
        "color": "#ffffff",
        "background_color": "#000000",
        "valign": "bottom",
        "align": "center",
        "y_offset": -12
      }],
      "imagemagick_stack": "v3"
    },
    "attributed": {
      "robot": "/meta/write",
      "use": "watermarked",
      "result": true,
      "data_to_write": {
        "copyright": "Copyright ${fields.rights_holder}"
      },
      "ffmpeg_stack": "v7"
    }
  }
}

Supply a reviewed rights_holder field when creating the Assembly. Use a short, nonempty name that fits the image; do not accept arbitrary ownership claims from an untrusted upload form. The same field supplies the visible notice and metadata so they cannot drift independently.

Only attributed is marked as a result. Verify both the visible pixels and the final metadata before publishing; metadata support and preservation depend on the file format and later tools. Neither notice creates permission that you did not already have.

Conclusion

Use metadata for triage, evidence for rights decisions, and watermarks or attribution for approved publication. Keep these steps distinct and test the final exported file. Automation can make the workflow easier to operate, but it must not turn absent metadata or a search result into a copyright clearance decision.