Summary of Content

PXV-ishii Tangible Bits: Beyond Pixels

This paper discusses a model of TUI, key properties, genres, applications, and summarizes the contributions made by the Tangible Media Group and other researchers since the publication of the first Tangible Bits

• 
  
  Tangible User Interfaces (TUIs) aim to take advantage of haptic interaction skills, which is a significantly different approach from GUI


• 
  
  TUI makes digital information directly manipulatable with our hands


• 
  
  Urp (Urban Planning Workbench)


• 
  
  In Urp, physical models of buildings are used as tangible representations for digital models of the buildings. To change the location and orientation of buildings, users simply grab and move the physical model as opposed to pointing and dragging a graphical representation on a screen with a mouse. The physical forms of Urp's building models, and the information associated with their position and orientation upon the workbench represent and control the state of the urban simulation.


• 
  
  The physical artifacts also serve as controls for the underlying computational simulation (specifying the locations of objects). The specific physical embodiment allows a dual use in representing the digital model and allowing control of thedigital representation.

GUI The figure illustrates the current GUI paradigm in which generic input devices allow users to remotely interact with digital information. Using the metaphor of a seashore that separates the sea of bits from the land of atoms, the digital information is illustrated at the bottom of the water, and the mouse and screen are above sea level in the physical domain. Users interact with the remote controls, and ultimately experience an intangible, external representation of digital information (display pixels and sound).

TUI Tangible User Interface aims at a different direction from GUI by using tangible representations of information that also serve as the direct control mechanisms of the digital information. By representing information in both tangible and intangible forms, users can more directly control the underlying digital representation using their hands.

The study then overview 8 forms of TUI, but only one applies to the application “Orchestrate”.

Interactive Surfaces,Tabletop TUI - Digital Desk is the pioneering work in this genre, and a variety of tabletop TUIs were developed using multiple tangible artifacts within common frames of horizontal. One limitation of the above systems is the computer's inability to move objects on the interactive surfaces. To address this problem, the Actuated Workbench was designed to provide a hardware and software infrastructure for a computer to smoothly move objects on a table surface in two dimensions [34], providing an additional feedback loop for computer output, and helping to resolve inconsistencies that otherwise arise from the computer's inability to move objects on the table.

Although this study doesn’t have any direct relevance to the project, it is in interesting way to explore the future possibilities of TUI’s.

P177-lucchi

This paper establishes the differences between touch and tabletop tangible interfaces in the quest to find the perfect interface between computers and people. Although an interesting study, it deals with precise figures as regarding accuracy and completion time of individual actions in both on the interfaces. The studies carried out on this paper don’t posses much relevance towards the project besides the direction TUI’s are moving towards.

p253- nishino

This Paper discusses the method of camera based fiducial tracking based on new methods, unique to the topological design of the fiducial structure.

Matrix-Pattern and Pattern-Matching Approach

• The use of matrix-pattern to encode IDs can be frequently seen in fiducial tracking systems.


• CyberCode, ARToolkit Plus and ARTag are examples of previous studies

Fiducial Recognition

The first one is the black square in the center, which contains one white dot. This is used to obtain the rough angle of the fiducial in the video input image, using the vector from center of the minimum bounding box of the black square to the white circle. The black and white regions are surrounding this center black square. Each dot in those regions encodes a bit, 0 for a black dot and 1 for a white dot. These bits are sorted in a clockwise order, starting from the rough angle obtained from the center circle, to decode the ID of thefiducial. Decoding these examples in Figure 5 results in their unique IDs, 48115, 64407, 40879, from left to right. Notice all these fiducials has the same topological structure. In case of 16bit fiducial, there are only 17 different topological structures, since the number of black or white bit-encoding nodes can vary only between 0-16.

The average time cost measured of 1000 frames with 12 markers in input images with capture in 640x480 resolution. ReacTIVision also has several features to increase the robustness like frame equalizer, the use of information from the previous frame and so on. Taking these features into the evaluation can increase the time cost. To compare these two different systems, we excluded the time cost for reacTIVision features that our system has not implemented yet.

This paper helps the reader further understand how fiducial tracking works and incorporate the best methods in use to optimize the speed and efficiency of fiducial tracking.