From CS294-10 Visualization Fa08

Jump to: navigation, search



12/12 Meeting #12

Paper work!

12/9 Meeting #11

Weights modification, going over presentation, splitting up tasks for paper.

12/6 Meeting #10

Fleshing out the poster for presentation, continuing work on heuristic, adding features such as palette/color preferences.

12/5 Meeting #9

Working more on the heuristic, making it a bit more modular (for easier updates). Creating some sample applications, screenshots, and preparation for the poster.

12/3 Meeting #8

Some discussion regarding traversal through the RGB and HSV space; turns out it's a lot easier to use the HSV space for comparison and checking between neighboring values as opposed to numerical calculations in the RGB space. Arranged for future meetings and wrap-ups before end of the semester.

11/25 Meeting #7

Modification to the UI; a background color has been added (with future changes to the heuristic needed to take this into account). UI is pretty much completed at this point.

11/21 Meeting #6

11/18 Meeting #5

Presentation preparation before midpoint check.

11/14 Meeting #4

Coding session to get some of the underlying code underway. Each member working on their individual "focuses", with collaboration/thoughts from every group member regarding implementation and ideas.

11/7 Meeting #3

Quick get-together on how we want to implement our ideas. Creating a new library using Flare, with individual parts/functions assigned to each group member.

10/26 Meeting #2

Worked on presentation, looked for papers covering similar ideas. Sketched an idea of what we'd like to implement, and how.

10/22 Meeting #1

Whittled down a set of potential problems to work on to one. Explored different ideas like Spore character, IP space visualization, and automated color palette generation based on user input. Chose to work on palette generation and set the next meeting time.




Tory and Möller, Human Factors in Visualization Research

§3.1.2, titled "Encoding Data With Color" cites a few papers we will want to look into.

Visualization systems often encode ordinal and quantitative
data using intensity or color gradients. For example,
topographic maps often represent elevation using a color
scale and medical images use a gray-level or color gradient
to distinguish tissues with different properties. However,
not all mathematically linear gradients are perceptually
linear (e.g., neither the mathematically linear grayscale nor
the rainbow (hue) scale are perceptually linear). For this
reason, several perceptually linear gradients have been
developed, as described by Levkowitz and Herman [46]
and Rheingans [64]. Most of these gradients are based on
variations in color value and/or saturation. 

Similarly, many visualizations use colors to segregate or
highlight objects. For example, a medical visualization may
show different organs in different colors and an air traffic
control display may use color to highlight potential
collisions between aircraft. Choosing colors for such dis-
plays is not easy because not all colors are equally
distinguishable by observers. For this reason, Healey has
devel oped a procedure for designing set s of easily
distinguishable colors [25]. 

Bergman et al. generalized these ideas in a taxonomy
based on principles of perception, visualization tasks, and
data types [3]. Their taxonomy can be used to develop and
choose effective color scales for specific data types and goals.

[25] C.G. Healey, “Choosing Effective Colours for Data Visualization,” Proc. IEEE Visualization, pp. 263-270, 493, 1996.

[46] H. Levkowitz and G.T. Herman, “Color Scales for Image Data,” IEEE Computer Graphics and Applications, vol. 12, no. 1, pp. 72-80, 1992.

[64] P. Rheingans, “Task-Based Color Scale Design,” Proc. Applied Image and Pattern Recognition, pp. 35-43, 1999.

[add comment]
Personal tools