Photographers and agencies are concerned about unauthorized image uses and the best ways to protect their copyright. The focus is often on catching the criminal—finding the individual who has made an unauthorized use—but there are two aspects to copyright protection that need to be considered. Much more attention should be paid to making it easy for individuals who find an image they want to use—whether on the Internet or in a printed publication—to locate the image owner and determine if a fee must be paid for the rights to use that image.
There are technological solutions, but a lot depends on motivation. It gets down to whether you believe that the first motivation of any individual is to steal whenever they think they can get away with it—or if you believe most people would prefer to be honest, if it is not overly burdensome to determine if the image owner wants a fee for use of an image and if that fee is reasonable. I believe most people who want to use images fall into the latter category.
It is important to make this distinction, because it affects how technology addressing this such situations should be designed. If we are trying to catch infringers, then we must first register every image we create, develop systems to identify infringement, hire lawyers, involve the courts and spend significant amounts of money in an attempt to collect damages. Even when all this is done correctly, image creators often end up spending more on the process than they recover. Identified infringers often have no resources, making it impossible to successfully claim damages. In such cases, all that earlier effort is wasted.
To make it possible for a potential user to locate the creator before using the image, all that is needed is a central database that contains identifying information of most of the images for which creators expect to receive royalties. A potential customer may then either upload a digital file of the image or scan an image found in a printed publication and upload a file of that scan. Once this information is received, the database is searched for an image or images with an exact fingerprint, and a message is sent to the customer listing the names and contact information of all sources the computer believes have an image that matches the one in which the customer is interested. The customer may then contact any one of those sources to negotiate usage rights. Given that many images are represented by a number of distributors on a non-exclusive basis, it is possible that many distributors and the copyright holder himself would be authorized to negotiate rights to use any particular image.
Even if the response takes a few hours or a day, it would still be an invaluable service. The Copyright Registry is attempting to accomplish this on a large scale. In contrast, PicScout and Idée’s image-to-image search engine TinEye are focused on identifying infringers and appear to be making no efforts in the direction of identifying owners.
Given the estimates of unauthorized uses on the Internet alone, I believe image creators would earn much more in the long run by focusing on the second group of image users rather than the first. By adopting a strategy that focuses on making it easier for customers to be honest, rather than on prosecuting infringers, image creators and distributors would certainly build much better long-term relationships with buyers.
To create an image fingerprint, the computer looks at an image, analyzes it in minute detail and then creates a unique alpha-numeric identifier (not a visual copy) of the image. These numbers are not a randomly assigned set; rather, they describe a specific image in the ways it is distinct from every other image that has or will be created—just as each human fingerprint is unique. Millions of image fingerprints can be stored in a small fraction of the space needed to store the images they represent.
PicScout, Idée and The Copyright Registry likely use a combination of visual search and some form of fingerprinting technology to match images, although the first two emphasize the visual search aspect and The Copyright Registry emphasizes fingerprinting. The differences among the companies lie in the costs in both time and money to participate and the program emphasis— catching infringers or helping customers find creators.