Compression reduces file size, but standard methods leave your image’s digital fingerprint intact — meaning reverse-image search still works against you. The answer: compress your image first, then apply protection that strips metadata, shifts colour channels, crops edges and adds a watermark. This makes the reverse-image fingerprint unrecognisable to Google Lens and TinEye, whilst keeping the visual quality useful for your client presentation. Every image is processed in your browser using the Canvas API — nothing uploads to the internet.
Why standard image compression doesn’t protect your work
Compression alone shrinks file size but does not obscure what an image is. A JPEG or PNG compressed to 200 kilobytes still carries metadata — EXIF data, colour profile, embedded software info — and more importantly, the same visual fingerprint. That fingerprint is what reverse-image search tools read. When a client shares a mood board containing an interior photograph, Google Lens can still match that image to the supplier’s website, even if the file size is half the original. The supplier then pitches directly to your client, and you lose the commission.
Interior designers, architects and specifiers face this problem daily. You spend hours curating mood boards and presentations. A single unprotected image — even a compressed one — can be reverse-searched by your client, who then bypasses you entirely to contact the manufacturer. Compression is not a defence; it is a false sense of security.
What does real image protection actually do to a compressed photo?
Protection that works against reverse-image search does four things simultaneously. First, it strips all metadata from the file — EXIF data, colour profiles, embedded authorship tags and timestamps all go. Second, it shifts the colour channels slightly, so the pixel values no longer match the original. Third, it crops a small margin from all four edges, changing the aspect ratio fractionally. Fourth, it tiles a watermark across the image in a pattern that breaks the visual continuity a search engine uses to identify the source.
Together, these changes make the image unrecognisable to reverse-image search algorithms. Google Lens looks for a visual and metadata match. When you remove the metadata, alter the colour space, crop the boundaries and add the watermark pattern, the fingerprint is no longer readable to the algorithm. The image remains useful and visually clear for your client presentation — the changes are subtle enough that the mood board still communicates the design intent — but the reverse-image match fails.
Can you compress and protect in the same step?
Yes. The practical workflow is: export your image at the size and compression level you need for your presentation, then apply protection before sharing. You do not need to maintain two versions. If your mood board needs images at 1200 pixels wide and 100 kilobytes per file, compress to those specs first. Then apply the protection layer — metadata stripping, colour shift, crop and watermark — which adds negligible file size and takes seconds in your browser.
The protection happens entirely in your browser using the Canvas API. No image is uploaded to a server, so your work never leaves your machine or your client’s email. This is important: you retain privacy and control. The file you download is the protected version, ready to share. There is no back-end processing, no cloud upload, no waiting for a server to return your image.
How do you know the protection is actually working?
The test is direct: take a protected image, open Google Lens or TinEye, and upload it. The search will fail to find the original source. Try this yourself with an unprotected mood board image, and you will see the difference immediately. Reverse-image search will return the supplier’s product page. Now protect that same image and try again — the fingerprint is broken, and the search fails.
This is not theoretical. Interior designers, architects and specifiers have validated this in their own workflows. The protection does not make images invisible (they are still visible to humans), but it makes them invisible to reverse-image algorithms. Your client can still see the mood board and understand the design direction. A third party with access to that mood board cannot easily find the supplier by searching the image. That is the point.
What are the practical limits of image protection?
Image protection is a fingerprint defence, not a legal one. It defeats automated reverse-image search, but it is not a copyright claim, a legal contract or a guarantee against all future forms of image theft. If someone crops your protected image very aggressively, removes large sections, or manually redraws key elements, they may be able to defeat the protection. If a person downloads your image and manually recreates it in different software, no protection can stop that. Image protection is a practical tool that stops casual supplier discovery and automated matching. It is not a legal remedy.
Also understand that protection is only useful if you apply it before sharing. Once an unprotected image is live on a website or in an email thread, it is too late. The original fingerprint exists in the wild. Protection protects forward — it stops new versions of the image from being easily searchable. It does not delete past versions that are already unprotected and publicly accessible.
When should you actually protect your images?
Protect any image that contains a product, finish, material or detail you do not want your client to source independently. If you are sharing mood boards with clients, protect them. If you are sending interior photography to a prospective buyer, protect it. If you are sharing material samples, finishes or supplier recommendations in a presentation, protect those images. The moment you send an image to anyone outside your immediate team, consider whether reverse-image search is a risk.
You do not need to protect every image. Photographs you have taken yourself, images in your portfolio that already bear your copyright or watermark, and generic reference images carry lower risk. But when you are assembling curated mood boards for client pitches — and especially when those boards contain commercially valuable supplier recommendations — protection is a straightforward investment. It costs nothing and takes seconds per image. The alternative is losing commissions to clients who bypassed you via reverse image search.