In all the discussion regarding the current Getty Images/Google Apps situation, a subsidiary of Getty Images, PicScout and their ImageIRC program has come up several times. People are wondering whether this has anything to do with the Google licensing scheme, that PicScout is, in some way, protecting our IP by their “fingerprinting” of images, or that Getty is going to use the ImageIRC program as some way to go after people who use content outside of the Apps program, which, we’ve been told, would be against the invisible licensing terms accompanying the content.
The following is all my understanding of how the programs work.
Now, the PicScout site itself says:
PicScout ImageIRC™ is the world’s largest index of image fingerprints and metadata from known copyright owners.
First, let’s define the “fingerprinting” of images. Some people are looking at “fingerprinting” as PicScout inserting some kind of tracking data into the image concerned. This is what DigiMarc, for example, is, but it isn’t what PicScout is. DigiMarc describes itself as:
Digimarc® for Images allows you to embed imperceptible, persistent digital watermarks into your images to communicate ownership and other information — wherever the images travel across the Internet.
PicScout is not a proactive action, but a reactive action. PicScout does not modify the content to make it trackable. It indexes available (or uploaded) content into its database, much like Google Images, TinEye, SpiderPic, etc. How it works:
- You send us your image files
- We fingerprint them and add them to our database of images
- We search the Web on an ongoing basis for images used in commercial and editorial websites
- We compare the images we found to yours
Proprietary ImageTracker methodology identifies image content based on digital fingerprints derived from algorithmic-based characteristics of the content. The highly accurate approach is independent of metadata, watermarks or file hashes and not affected by alterations that may have occurred in applications of an image. PicScout has developed the most widely used and best approach for identifying image content anywhere and in any context on the Web.
Photoshelter works (or did work at one time) with PicScout and this is how they describe the process:
In order for PicScout to fingerprint your images, PhotoShelter securely sends a low-resolution copy of the image to them. Then PicScout fingerprints your images using an advanced image recognition algorithm that identifies unique patterns within the image. Unlike a watermark, this image fingerprint is derived from the image and therefore cannot be erased, modified, or edited. The current ImageExchange technology allows PicScout to identify images in a near-exact match as they appear on the Internet.
Each image fingerprint is associated with its unique metadata, which we send along with the image. At present, this metadata includes image owner, license type, and image URL. The image is not altered by PicScout in any way.
Using the word “fingerprint” makes it sound more sexy and mysterious then just “indexing an image”.
ImageExchange is a PicScout browser plugin that helps those wishing to properly license imagery easily use the “fingerprinted” database to find the owner and/or stock site where the content is available. When you find an image on a page you wish to license, you can use the plugin to retrieve the data from PicScout to help you legally use the image.
Using the “fingerprint” database to power their ImageExchange plugin is different from the “Post-Usage Billing” program they have come up with.
Today, we are excited to announce the first implementation of the ImageIRC Post-Usage-Billing Service. The ImageIRC Post-Usage Billing Service is a service that facilitates credit and compensation of professional imagery shared on social image-sharing platforms. This new solution will ensure that professional imagery shared by platform users is properly credited and licensed.
Essentially, when images are uploaded to participating social media platforms, the platform uses an API to query the ImageIRC database to see if the image can be found. If so, the social media platform is charged some amount for using the material, and, theoretically, that money gets back to the contributor as well as the agency.
To tell you a bit more about how the ImageIRC Post-Usage Billing Service works, social image sharing platforms begin by integrating with the ImageIRC API. When a user shares an image, the ImageIRC API is pinged to determine if the image can be identified and licensed. When possible, the ImageIRC grants a license for the image at the time it is shared. Content licensed through the ImageIRC Post-Usage Billing service will display the licensor’s credit and caption within the image.
So, theoretically, if Facebook had integrated this product with their platform, and someone uploaded an image from the Getty library that had been “fingerprinted”, and the system was able to realize that, FB would be charged in some way for use of the image.
Neither the ImageIRC program, nor the Post-Usage program, are inherently evil or bad for photographers. They just don’t really have anything to do with the Google/Getty situation as it is, as far as I can tell.
Now, if you were a stock image buyer/user, and a bit paranoid, and you are afraid of getting extortion letters from Getty Images for improperly using content, I would still shy away from Google Apps, just in case they have something nefarious up their sleeve.