Re: 90 Rotation in one buffer only
- From: "Jens Dierks" <me@xxxxxxxxxxx>
- Date: Fri, 25 Nov 2005 11:51:44 +0100
Naresh wrote:
> Hii All,
> what could be the BEST Memory Optimal way of rotating an image or
> matrix by 90
> degrees using input buffer only? i.e. Input buffer only is the output
> memory buffer.
>
> Say, we can use small amount of memory locations (say 30-40 bytes extra
> considering 1 pixel/byte) for temporary storage of any block/blocks in
> the image but not the whole of image.
As said, its not so easy done with rectangular Pictures.
There is no simple solution with just a tmp buffer because it
depends on the relation of width and height how often the
loop P[x,y]=P[x',y'] will be closed and a new loop starts.
The Problem is how to find the new starting point without
marking all changed Pixels.
Maybe there is a mathematical tricky way like fft transform,
i just dont know it.
Greets
Jens
.
- Follow-Ups:
- Re: 90 Rotation in one buffer only
- From: Naresh
- Re: 90 Rotation in one buffer only
- References:
- 90 Rotation in one buffer only
- From: Naresh
- 90 Rotation in one buffer only
- Prev by Date: Re: Object Tracking in 3D Space
- Next by Date: Re: 90 Rotation in one buffer only
- Previous by thread: Re: 90 Rotation in one buffer only
- Next by thread: Re: 90 Rotation in one buffer only
- Index(es):
Relevant Pages
|