IMATAG: Protecting Digital Assets Better
When it comes to businesses, just like regular data, visual contents also play a critical role. Images are an integral part of companies’ commercial strategy and a visual revealing information about a new product or unveiling a new campaign has a considerable value, namely the cost of losing control of it. However, visual content is not data like the others, it requires specific tools to protect it. Companies selling products or having suffered leaks of sensitive visual content wish to know the origin of these leaks to prevent them from happening again. It can be photos, illustrations, or videos. They often represent future products, but they can also be exclusive content or confidential visual information. Their particularity is that they are not intended to remain locked in a virtual safe, they are shared between employees, partners, suppliers, and market research panels. These images are often “protected” by an NDA, sometimes accompanied by an embargo period. The recipients are multiple and can be internal, or external, the exchanges are done by standard channels (mail, drive, sharing a link, download with login/password…). Imatag’s invisible watermark takes the security of visual content to the next level. “With Imatag, Forensic watermarking, historically reserved for multimedia industries (digital cinema and early release movies) to fight piracy, is now accessible to all industries whose success depends on controlling their images (still and video), thanks to our patented process,” explains Mathieu Desoubeaux, CEO, Imatag.
With IMATAG, visual content is marked invisibly with a unique identifier. Example: an image distributed to 5 partners will be marked with 5 different watermarks. If one of these images is disclosed outside the conditions initially agreed, the detection of the watermark it contains will reveal the initial recipient of this copy, and help investigate more deeply the stages of the leak.
When an image must be transmitted to several recipients to be tracked, or to know through which channel the leak occurs, a unique copy per recipient or channel is produced by the system, hiding in its pixels the code which will allow it to be identified. Unlike other solutions on the market, the Imatag watermark does not visibly alter the original image, and it remains detectable despite strong transformations such as cropping, resizing, strong compression (as often happens on social networks), and even screenshots. The procedure for detecting the origin of a leak is simple: the leaked image (even an extract) is submitted to the detection system. If the mark is found in the image, the corresponding key delivers the information initially entered during its marking (IDs of a campaign, a session, a recipient, a channel, etc.). Once this first source has been identified, the company can investigate how the leak could then have occurred. “Our system is patented, and issued by a renowned research institute. The embedding process maximizes the security, robustness, and quality of the watermark,” adds Desoubeaux.
There are very few offerings today when it comes to still image protection, and none have the robustness and imperceptibility of IMATAG. Forensic watermarking techniques invented for the multimedia production industry are adapted, neither in price nor technically, to the needs of brands. With IMATAG, it's the same solution for images and videos. Large companies from consumer electronics, automotive, entertainment, and luxury industries have chosen Imatag after conducting comparative tests. The invisible watermark works by modifying slightly the values of the image’s pixels according to a unique key and a perceptual mask to exploit the properties of the human visual system (areas of the image where the watermark modifications are stronger).
Since opening its doors in 2015, Imatag has been a pioneer in delivering cutting-edge image security solutions. The company has already made its presence on the international market, has partnerships with DAMs, and integrators, and contributes to initiatives like CAI (Adobe's Content Authentication initiative). “We believe simplicity and integration are important to our customers. So, we simplify the marking of other formats and types of sensitive content, and we integrate it with online content-sharing tools. Our customers would also like to have the tagging directly integrated into their own sharing tools, so this is where our roadmap is heading” concludes Desoubeaux.