New: this project is also covered on d-touch.org.

Visual markers, such as barcodes and 2D-barcodes, are graphic symbols designed to be easily recognised by machines, such as computer and mobile phones. Traditionally markers are not visually meaningful to people, and different markers of the same family are not easy to distinguish from one another by looking at them.

d-touch is a visual marker recognition system that supports markers that are both machine-readable and visually communicative to humans. d-touch allows users to create their own visual markers, controlling their aesthetic qualities and what they visually communicate to others.

By allowing the creation of markers that support interaction both visually and functionally, d-touch can enhance most applications normally supported by visual markers, including interactive guides, mobile service access, mobile games, interactive story telling systems and augmented reality applications that have broad visual appeal and are not constrained to ugly glyphs.

Set of 4 demo markers for business cards verso, designed by Tal Drori.

Markers that are both functional and visually expressive can be easily produced by a wide spectrum of users, without specific training. During a user study investigating marker creation pairs of novice users generated between 3 and 27 valid markers within one hour of being introduced to the system, demonstrating its flexibility and ease of use. Because d-touch markers are designable, end users and designers can consciously determine their look and feel, including the degree to which they are immediately recognizable as markers to be scanned. The design can range from icons that are obviously scannable (explicit) to ones that are hidden in the overall design and only accessible to a closed circle or upon closer look (ambiguous).

d-touch runs in real-time on both mobile phones and desktop computers and it is released under an open source license (GPL). The creation of markers which are both machine- and human- readable is possible because d-touch recognition is based on topological features of the markers rather than their geometry. Marker recognition is not based on shape, but on the relationship of dark and light regions. To increase computational efficiency, d-touch imposes constraints on the design of the markers in terms of the relationship of dark and light regions in the symbols. More information about topological recognition is available here. An additional benefit of topological recognition is that it allows to recognized multiple object at the same time with negligible performance degradation, for this reason it was used as an enabling technology for a number of tangible user interfaces, including initial reacTable* prototypes (later on the reacTable* team developed reacTIVision, its own specialized marker recognition system, based on the concepts introduced by d-touch).

Part of a set of markers designed to add multimedia content to a book, by Giuseppe Costanza.

More information about d-touch markers is available in the Designable Visual Markerspaper. Earlier papers describe other technical aspects of the system.