[Date Index]
[Thread Index]
[Author Index]
Re: Image[], Graphics[Raster[]]
*To*: mathgroup at smc.vnet.net
*Subject*: [mg100284] Re: Image[], Graphics[Raster[]]
*From*: Jens-Peer Kuska <kuska at informatik.uni-leipzig.de>
*Date*: Sun, 31 May 2009 06:33:01 -0400 (EDT)
*References*: <6147782.1243412890107.JavaMail.root@n11> <200905280829.EAA14200@smc.vnet.net> <gvqpu5$ck0$1@smc.vnet.net>
Hi,
>
> 1. Is Image[] any different than Graphics[Raster[]] that I'm more
> accustomed to?
No it has a fixed data range, i.e., "Byte" is [0,255], "Real" [0,1]
while Graphics[Raster[]] can handle undestructive every thing.
>
> 2. If not, what was the point of introducing Image[]? Was it any effort to
> provide a more streamlined set of commands all revolving around Image[],
> e.g., ImageRotate[], etc.?
The reason was a faster display of the bitmap. With fixed data range
the FE don't have to do a scaling of every pixel.
>
>
> 3. Image[] somehow upsets my unified view of Mathematica. That is, I like
> to try to "think" like Mathematica, and so I understand that all commands
> like Plot[], ListPlot[], etc., reduce to a Graphics[] object in
> FullForm[]. Similarly, I can build my own type of plotting function with
> Graphics[]. Most importantly, I can combine many different types of
> objects into a single Graphics[]. Am I correct that Image[] represents
> then a completely new display object? If so, is there a way, for example,
> to have an Image[] and then draw across it with an Epilog rule like you
> can do with Graphics[]? It seems a major setback if Image[] and Graphics[]
> cannot be readily mixed.
>
Oh mixing is easy:
img = Import["http://www.stephenwolfram.com/img/home/sw-portrait.jpg";]
Graphics[
{Image`ToGraphicsRaster[img][[1]],
Circle[{110.5, 110.5}, 10]}]
draw a circle on the image.
> Anyone out there to make the case that Image[] represents an important new
> capability?
No, it make many things more complicated, and a lot of image
processing tasks impossible -- registration for example ..
I see it as a little worse than the usual "new command that
> does same thing old stuff did" because it seems to lose the capability to
> mix Graphics[] and Image[].
>
> Please inform me if I'm completely off base. As I said, I'm just digging
> into Image[] now for the first time, and I'm a bit confused about possible
> merits---please do inform!
There is no merit, it make many things worse than the original
Graphics[Raster[]]
Regards
Jens
Prev by Date:
**Re: problem writing debugging utility function**
Next by Date:
**Re: Image[], Graphics[Raster[]]**
Previous by thread:
**Image[], Graphics[Raster[]]**
Next by thread:
**Re: Image[], Graphics[Raster[]]**
| |