Images, Resolutions, and IAO
2 views
Skip to first unread message
Albert Goldfain
Feb 10, 2009, 4:08:32 PM2/10/09
to rj...@buffalo.edu, informatio...@googlegroups.com
Ron Rudnicki brought up an interesting issue regarding images in IAO...so I thought I would copy him and the IAO on this.
Currently, images are classified as generically dependent continuants (GDC) in IAO:
"An image is an information content entity which provides a visual representation of for the result of a measurement."
The issue is that many would like to say that images have resolutions (e.g., my laptop's background image has a resolution of 1600 x 1200 pixels) as a quality, but as Barry pointed out, a resolution would be a further GDC (of the GCD image) and any case of this "double dependence" can be reduced to a single dependence.
My solution (allowing us to keep images where they are in the ontology) would be to say that no matter the way we speak of images, it is not them but certain types of their renderings which have a resolution.
A resolution is a quality of an image rendered in a pixelated format. This is why the measurement of resolution is given in terms of the unit pixel. A pixel stands for "picture element" (not many people mention this...there are also things in image processing called texels "texture elements"). A pixel is a pixel of a particular image rendering on the basis of:
(1)being of a unit size in the rendering format
(2)some mereological facts: a pixel has RGB value (255, 0, 0) if the corresponding thing the rendering is an image of has a part that reflects red (under the appropriate lighting etc.)
I am perfectly willing to accept some strange "renderings in pixelated format" as long as only a subset of these are images qua information artifacts: jpeg renderings (digital pixel), mosaics (tile pixel), and tapestries (knit pixels) are all pixelated renderings of images, but Barry's example of linen sheets (which have a thread count) are not even though they *are* pixelated. The former three are image renderings because their pixels are informational (i.e., carrying information *about* their content).
Also, there is something to be said about the connection between the pixelation of certain image renderings and their being digital entities.
Any thoughts?
-Albert
Currently, images are classified as generically dependent continuants (GDC) in IAO:
"An image is an information content entity which provides a visual representation of for the result of a measurement."
The issue is that many would like to say that images have resolutions (e.g., my laptop's background image has a resolution of 1600 x 1200 pixels) as a quality, but as Barry pointed out, a resolution would be a further GDC (of the GCD image) and any case of this "double dependence" can be reduced to a single dependence.
My solution (allowing us to keep images where they are in the ontology) would be to say that no matter the way we speak of images, it is not them but certain types of their renderings which have a resolution.
A resolution is a quality of an image rendered in a pixelated format. This is why the measurement of resolution is given in terms of the unit pixel. A pixel stands for "picture element" (not many people mention this...there are also things in image processing called texels "texture elements"). A pixel is a pixel of a particular image rendering on the basis of:
(1)being of a unit size in the rendering format
(2)some mereological facts: a pixel has RGB value (255, 0, 0) if the corresponding thing the rendering is an image of has a part that reflects red (under the appropriate lighting etc.)
I am perfectly willing to accept some strange "renderings in pixelated format" as long as only a subset of these are images qua information artifacts: jpeg renderings (digital pixel), mosaics (tile pixel), and tapestries (knit pixels) are all pixelated renderings of images, but Barry's example of linen sheets (which have a thread count) are not even though they *are* pixelated. The former three are image renderings because their pixels are informational (i.e., carrying information *about* their content).
Also, there is something to be said about the connection between the pixelation of certain image renderings and their being digital entities.
Any thoughts?
-Albert
Reply all
Reply to author
Forward
0 new messages