Kevin: Journals

Hey guys, as promised, summarized the Journals I've read into point form and placed them below. Note that the ratings next to the titles are usefulness for our project.

In my opinion, we should definitely mention the Advantages, Limitations, and Uses for StitchRV (found in Wang's Journal).

StitchRV is a software application which allows for multiple camera inputs, meaning we can increase the reactive table's range, make the fiducials smaller, or get the cameras closer to the fiducials (which is useful as it limits the bulkiness of the rig).

StitchRV is currently limited due to the restricted number of cameras, which is dependent on processing power of the workstation. (In tests using a Macbook Pro, dualcore 2.4GHz Processor and 2Gb Memory, the system could only handle 3 cameras)

Another (current) limitation is the fact that changing settings (such as number of cameras used) requires altering the source code (C++), however this shouldn't be an issue for seasoned programmers.

Also note how the second journal (Out of the Box) has been rated as a (1/5). This is not due to misleading content or lack of interesting points, however the entire journal deals with a single case study of a reactive Lego table in a toy store. Table would display 3D images of toys once their box was placed on the sensor part of the table.

All in all a good read, but I personally did not find it applicable to our cause.

Anyway, summaries or the 4 read journals may be found below, enjoy.

P6- Vandenhoven: Tangible Play (4/5)

· JJournal deals with tangible games, ideal for our scenario.

· “They [people] like game play for a variety of reasons: as a pastime, as a personal challenge, to build skills, to interact with others, or simply for fun” Quote from text, may be useful as an impact line at bottom of slide.

· Journal also specifies that some gamers prefer board games over newer games to due socialisation aspect: I do not fully agree with this, as certain modern games allow for socialising (via online play, two player mode, or multiple input such as Reactable)

· Vandenhoven in fact went on to mention the conceot of sitting around a table to play digital games (via embedded touch displays)

· The ‘Tangible Play’ workshop “brings together researchers and practitioners”, and deals with topics such as marketability of tangible games.

· Tangible Play is a one day workshop, in which a “guest speaker from Philips Research ... will provide an industry perspective on tabletop game design.”

P61- Nielsen: Out of the Box (1/5)

Deals with children’s physical toy ’marketing’ by bringing them to life at a toy store, via 3D game engine and reacTIVision game table.

· LEGO toys are designed using a 3D program. That same design tool is eventually used to create the physical moulds.

· To get point 1 to work, a reacTIVision table was constructed, with a camera located underneath the Perspex glass table, and a monitor mounted on a nearby wall.

Toy boxes all carried fiducials underneath, which when placed on the reacTIVision table; prompt the system to show a 3D model of the toys on the screen.

It is also good to point out that when they where constructing the table, they had their target audience (6-16 year old children) in mind, designing the table’s height accordingly.

· Similar reacTIVision tables include:

o “Tangible Programming exhibit at the Boston Museum of Science”

o “Battleboard 3D [which] is a mixed-reality chess-like board game”

o Storysurfer which is a “combined surface (floor: top-projected, table: back-projected) installation, which allows children to browse and select books in a library”

· The majority of this paper deals with how the authors constructed their toy store table, where they positioned it, and contains a study of who made use of it during the span of the study (gender and age group studies). As interesting as this study was, most of the data is inapplicable to our group’s scenario.

P63-Van Dam: Post-WIMP User Interfaces (4/5)

· Four styles of interface throughout time:

o 1950s-1960s: Batch processing (Punch Card Input, Line-Printer Output): No user interface due to lack of interaction with system.

o 1960s-1980s: Basic Command Line Interfaces, via “alphanumeric displays”

o 1970s: Launch of “Raster graphics-based networked workstations”, along with “point-and-click WIMP GUIs”

o 1990s onwards: Post-WIMP user Interfaces

· Van Dam considers post-WIMP to be interfaces which do not make use of “menus, forms, or toolbars, but rely on ... gesture and speech recognition”

· “the most important predictor of an application’s success was its ease of use (how “user-friendly” it is), both for novices and experienced users” Quote from text, may be useful as an impact line.

· Mentions concept of butler style interface, where workstation knows what user requires without direct user input (can be done via facial expression/hand gestures)

· Drawbacks of WIMP GUIs

o As application complexity increases, it becomes harder to learn the interface, due to an increased number of widgets and features.

o Users spend too much time “manipulating the interface” to get to the point where they wish to get to (due to point-and-click methodology)

o Mouse and keyboard input not available to all users (either because they don’t find controlling a mouse to be natural, or due to injury related issues/disability.)

o “WIMP Interfaces do not take advantage of speech, hearing and touch”

· The Future: “Hugely powerful ubiquitous computers in many different form factors”

o “Wearable Computers” (6^th Sense)

o “Whiteboard sized or wall sized displays” (already in existence)

o “Lightweight, minimally intrusive head-mounted displays” (Virtual/Augmented Reality)

o Future display resolutions will be far better then the current 70-100dpi, which Van Dam considers inadequate.

P287-Wang: StitchRV: Multi-Camera Fiducial Tracking (5/5)

· StitchRV: fiducial and touch-tracking engine based on reacTIVision.

· Combines video input from multiple cameras at the same time

· Single camera setups are limited to the field of view of their single camera. Adding more cameras to the rig increases the field size, and reduces the chance of obstructions.

· Hardware: Currently designed to work with a Playstation Eye which delivers video via USB at 60frames per second. (640x480 resolution)

· Advantages of StitchRV:

o Requires little equipment

o Enables object-tracking

o Multiple cameras allow for higher resolution fiducial and touch tracking

o Through point 3, allows fiducials to be smaller (allowing for a more densely packed surface)

o Multiple cameras makes it possible to have less distance between cameras and surface (as surface range not limited by a single camera’s line of sight)...

· Current Limitations:

o Only supports two camera feeds (at the moment)

o Source Code customization (openFrameworks / C++) may be intimidating to some researchers

o Currently not a viable solution for researchers with no programming experience

o Limited by processing power

· Future work includes User Friendly Interface (to replace requirement for source code editing)