AN ANALYSIS OF INTELLECTUAL PROPERTY PROTECTION TECHNIQUES FOR DIGITAL CARTOGRAPHIC DATA ADAPTED FROM CARTOGRAPHIC COPYRIGHT TRAPS USED ON PRINTED MAPS

M. Rice

NCGIA Postdoctoral Fellow, UCSB Department of Geography, Santa Barbara, CA, USA

rice@geog.ucsb.edu

For millennia, knowledge about the Earth has been organized, presented, and described through maps. The verbal, textual, and graphical information used in creating maps has always been a valuable commodity. Many approaches for protecting the intellectual property of maps have developed as a direct outgrowth of marketing and commerce opportunities, particularly during the exploration and colonization of distant lands by European powers. This research explores that history, with an emphasis on ways that artificial artifacts — cartographic traps — have been used on paper maps as a means of identification and intellectual property control. Cartographic traps are effective because they are hard to detect and remove, and have been employed by a variety of different map publishers, both historical and contemporary. Contemporary examples of cartographic traps and a short analysis of techniques is presented to document their use.

Maps are now stored primarily in digital format. This research looks at the cartographic trap on printed maps, and explores how similar concepts could be instituted with digital geospatial data. Vector geospatial data are often stored as coordinates using double precision variables (64 bits), which can define position on globe to the micron level (10^-6 meters). Many of the 64 bits are not used for defining position but are simply an artifact of computer programming. An experiment conducted by the author looks at simple strategies for embedding an identifier in vector coordinate digits, and tests the degradation of the identifier through the repetitive cartographic projection transformations. Some cartographic projection transformations, such as the Albers Conic projection and the Sinusoidal projection, show very stable behavior with regard to numerical stability and information loss. Other projection transformations, such as the Bonne projection and the Cassini projection, as well as compromise projections such as the Robinson projection, show numerical instability and a level of information loss which would prohibit large-scale embedding of identifiers using simple techniques. The results of this experiment are discussed in relation to current research in geospatial data sharing and relevant suggestions by other authors that technology such as digital watermarking and steganography should be employed for protecting the intellectual property rights of digital cartographic data.