Owners of digital media (images) are cautious about making their work available to the public due to the risk of illicit copying and distribution. Withholding images protects the author's work but also prevents the author from gaining interest and recognition from the public. Making samples of the work available raises public awareness but also makes that work subject to theft. If the theft of an image is suspected, then how can one locate the image over a distributed system such as the Internet? We consider several methods for image recognition: methods applied by image database systems, digital watermarking techniques, and an alternative method of image recognition. In doing so, we provide classifications of image database systems and digital watermarking methods. We then introduce an alternative method and illustrate how it fills in the void of image recognition between image databases and digital watermarks.

Click Here to download this article