Imatag protects images with invisible digital watermarking.
But what does this mean? Why is it different from visible watermarks? Is it just for copyright protection? Find out the answers with traditional examples and more innovative cases.
Quick read :
- What are the main differences between visible and invisible watermarking?
- What are the mainstream uses of digital watermarks?
- 6 innovative watermarking use cases for your business
- 6 tips to chose your digital watermark solution
The great confusion : visible and invisible watermarking
The term "Digital watermarking" was coined by the research labs, which invented this new technology.
Yet, in digital photography, this wording brings a lot of confusion since photographers have been using "watermarking" for a long time. These two "watermarking" are not the same.
A digital watermark is invisible
Digital watermarking provided by Imatag adds an invisible watermark to the picture. The pixel values are changed in order to convey a secret message, but these modifications are subtle and perceptually shaped. The human eye cannot see the difference between the watermarked and the original image.
This is a major difference with visible watermark which is an overlay superposed on top of the picture and meant to be seen. To avoid the confusion, we should speak about invisible digital watermarking.
A digital watermark embeds a code
Digital watermarking embeds a secret message within the image (the technical term is steganography). This message is binary and can be decoded by a computer.
This is the second major difference with visible watermarks which are meant for human readers. This code can be unique (content ID) to help identify different copies of a single image.
A digital watermark is more robust
In some languages (in french for instance), this technology is also called "digital tattoo". This conveys the idea of robustness. Even if the image has been edited (JPEG compression, cropping, scaling, color editing, ...), the computer can find back the embedded message. This message sticks to the image as a tattoo on the skin. You can play with our demo tool to challenge Imatag's watermark robustness, or read test reports from PetaPixel or DPReview.
On the contrary, visible watermark, though its visibility is meant to deter people from stealing photos, is easily removed. The simplest way is to crop the image when the watermark is placed as a signature. Regarding watermarks spread all over the image, they can be deleted by artificial intelligence, as detailed in our article about the algorithm used by Google. Most photographers are aware of this weakness, this explains why visible watermark is the third tip they use to protect their images, and not the first one (read Professionals’ tips on how to protect your images from online theft).
6 mainstream uses of digital watermarks
Invisible digital watermarking is not new and it is everywhere.
The first research works date back to 25 years ago. This technology is now well spread because it can be applied to many types of content (still images, video, audio, source code, maps, 3D model, DNA sequence, chemical compound, ...) and its use cases are legions. You probably have already seen or heard watermarked content without knowing it. This is just the point : digital watermarking is invisible, or inaudible, imperceptible ...
Some applications of this technology are the following :
1- Audience monitoring
In Europe and in the USA, the audio of TV channels is watermarked. The embedded message encodes the name of the channel.
Panelists have at home a device capturing the sound in their TV room. It decodes the message and sends it to the monitoring companies via Internet. They can measure the audience and market share live.
2- Content integrity
In Africa, the TV ads are typically long spots. They are broadcasted through many local channels which sometimes shorten the ads to increase their number and whence the revenues. The embedded message is the identifier of the TV spot followed by a time-code. Some companies have platform listening to all TV channels 24hours a day and check that any single second of a TV spot is duly broadcasted.
3- Content protection
The audio of movies are watermarked in theaters and on Blu-Ray discs. The decoder is embedded in a Blu Ray disc player.
If it detects a watermark while playing a content in the clear, it concludes that this content is indeed a pirated copy (either a camcord of a theater projection or a ripped Blu Ray) and it stops the playback.
Videos in movie theaters are also watermarked. The embedded message is the date of the projection and also the ID of the beamer. That way, "Hollywood" knows in which theater a pirate has camcorded a movie and puts the pressure on the owner of the theater.
5- User tracing
Websites selling multimedia content on demand (movies, songs, cartoons, e-books) send the client a personalised copy of the content. The ID of the client is embedded so that dishonest people illegally sharing their versions (like on P2P networks) are identified and black-listed (or sued).
6- Packaging identification
Some companies use digital watermarking as a replacement of the bar code on packaging.
6 innovative watermarking use cases for your business
In short : digital watermarking is invisible, not new and already used in many applications. But today, invisible digital watermarking also has new helpful applications for your business. Whether you sell images or your business relies on them, reasons for checking what they become once you released them are many.
Audience monitoring (brands, PRs, corporate, e-commerce)
Even if you are not a broadcaster, you probably use images for your own business, for communication, for marketing. Monitoring just your own website's audience or following your performance on social networks is not exhaustive, considering that your visual content can be reused outside these walled gardens.
Search engines limitations : an easy way to find your images online is to use online reverse search engines like Google Image or Tineye. But as their technology is based on similarity and not digital watermarking, it has at least 2 drawbacks : trade-off between false positive / false negative and multiple sources (for more details please download our white paper).
Text (metadata) limitations : you can also use text search to find images containing your credit or copyright in their metadata. But here again, some drawbacks will make it difficult :
- More than 97% of images online do not contain such information (most of them have no metadata at all), as revealed in our metadata studies (State of image metadata in 2018 and Luxury & Fashion: does branding matter more than product images?).
- Main search engines form the market do not use the metadata information in mages as a search key. Lately Google Image announced that they now display the IPTC field (read our article about metadata standards), but so far Imatag's search engine is the only one allowing to query by copyright.
Digital watermark advantage: monitoring your images online with digital watermarking is the only way to certify that images found online were initially sourced by you, and thus the only reliable way to script automated process like invoicing, claiming or sending a DMCA, building statistics and reports ...
You can check your images integrity to ensure that your brand chart or your reputation is respected, to detect counterfeit or grey market. See more details in our article 5 reasons why brands should monitor their images.
Forensics, user tracing
A premature release of an image to the public can have devastating consequences to some brands. Some leak-sensitive brands like smartphone manufacturers (see our case study) ask their contractors to add invisible watermarks in their sensitive content to prevent content leaks.
In other cases where exclusivity or scarcity makes the value of an image, the photographer or its agency can decide to end a contract with a client who has published a photo without authorisation.
Platforms like Youtube detect the use of media (mainly audio and video) to protect copyright holders subscribed to this service. To do so, a signature of these medias was recorded in the database. Similar solutions are proposed based on a blockchain, but with certain disadvantages (read also "About blockchain and decentralized copyright management of images"). Watermarking images is the the safest way to reverse seach a content-ID database due to the increased accuracy of match results.
Fact checking, fake news inquiry
The metadata accompanying the photos are mostly deleted when they are published (read our study on metadata in press). As a consequence, apart from copyright issues, it becomes impossible to know, when an image accompanies a news, if it corresponds well to the event (place, date, circumstances) mentioned by the article. Watermarks enable to maintain a link between a photography and its original source, with the same mechanism as content-id. Read also "A watermark to arm publishers against fake news".
Accurate reverse image search
Most mainstream reverse search engines rely on visual recognition algorithms, which accuracy cannot comply with decision-making automation. As a consequence, the cost of labor to check manually the reported matches is prohibitive for many services based on visual recognition.
But combined with web crawling and scrapping techniques, a digital watermarking solution provides the same service as a reverse image search engine, while offering enhanced reliability on matches.
Here is an example of use based on a patented watermarking solution, Imatag, to monitor usage of your images on the web, and for leaker tracing.
If you are using reverse search tools to track your images, read our article to learn more about reverse image search limitations, workarounds and alternatives.
6 tips to chose your digital watermark solution
If imperceptibility is key, if your images are to be compressed, if you need to identify different copies of a unique image, if your content (even samples) cannot be stored outside your own servers, if interoperability with your DAM is not an option, if you need to integrate the watermarking process into your existing workflow, make sure to check taht all these features are available and test them.
Here are some of Imatag’s differences with our competitors :
Robustness to compression
The jpeg compression, systematically used on all press sites and others, easily removes the watermarks (see "Are my images protected on Facebook and other social media sites?"), contrary to the Imatag watermark which resists to extreme compression rates (test it on "https://www.imatag.com/demo/watermarking/).
After zooming, some watermarks become noticeable under the form of a green "noise" crippling. Imatag’s noise is hidden in the grey levels close to high contrasts areas, a much more inconspicuous solution.
Imatag is able to add a different watermark to each copy of a single image. The watermark acts as a tracer, it can link each copy to data such as the channel used for its distribution, or date and name of the DAM user who downloaded it... Other digital watermarks will only apply the same pattern to all your images, or mark each copy of an image with the same watermark.
Unlike Imatag, competitors encryption is slow for voluminous images. Carefully evaluate the amount of images that will need to be daily processed to make sure your solution will be able to scale.
Our API is open to all customers from the cheapest plan to the highest. A plugin for LightRoom users is also available, allowing them to save on their own disk the watermarked photos at their original size and automatically trigger their monitoring online.
Sensible data storage
Imatag uses first a low definition copy of the original photograph to match it with the query image before identifying its code.
This also allows to track unmarked images as Google Reverse Image Search and Tineye do for example. If monitoring of your images is not required, but only authentication (for traitor tracing for example), Imatag does not need to store your content. Please contact our sales team for details.