Path: news.mathworks.com!not-for-mail
From: "Sebastian Bludau" <s.bludau@fz-juelich.de>
Newsgroups: comp.soft-sys.matlab
Subject: Reading images of size >1gb on 32bit System
Date: Fri, 16 May 2008 08:18:02 +0000 (UTC)
Organization: FZJ
Lines: 22
Message-ID: <g0jfvq$htm$1@fred.mathworks.com>
Reply-To: "Sebastian Bludau" <s.bludau@fz-juelich.de>
NNTP-Posting-Host: webapp-03-blr.mathworks.com
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 8bit
X-Trace: fred.mathworks.com 1210925882 18358 172.30.248.38 (16 May 2008 08:18:02 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Fri, 16 May 2008 08:18:02 +0000 (UTC)
X-Newsreader: MATLAB Central Newsreader 1388132
Xref: news.mathworks.com comp.soft-sys.matlab:468777


Hi,

i would like to read tif images with a possible file size
larger than 1gb. If i use feature('memstats') the largest
continuous memory block on my system is about 900mb.

Is there a easy way to read those images without having the
"out of memory" problem on my system.

Is it possible to resize the image while reading it, so that
the image fits into my memory? 

I thought about 
[...] = IMREAD(...,'PixelRegion',{ROWS, COLS})
Maybe i can read parts of the image, resize these parts and
put the resized part into a new matrix which should contain
the final rescaled image. (Hope there is a more easy way to
handle my images)

Thx for any ideas

Sebastian