Copy Link
Add to Bookmark
Report
VISION-LIST Digest 1990 08 06
Vision-List Digest Mon Aug 06 17:24:54 PDT 90
- Send submissions to Vision-List@ADS.COM
- Send requests for list membership to Vision-List-Request@ADS.COM
Today's Topics:
SRI testbed to raster format..
Re: Displaying 24bit/pix images on 8bit/pix device
Re: Converting 24 bit images to 8 bit.
Re: Displaying 24bit/pix images on 8bit/pix device
Re: Displaying 24bit/pix images on 8bit/pix device
Feature extraction
Landscaping data
Change in CVGIP Journal
----------------------------------------------------------------------
Date: Fri, 3 Aug 90 14:05:11 PDT
From: venkat%prospero.usc.edu@usc.edu (V. Venkateswar)
Subject: SRI testbed to raster format..
I have some images which are in the SRI testbed format. What exactly is
this format? I am trying to figure out a way to convert this to ordinary
raster format. Thanks for any information.
V. Venkat (venkat@prospero.usc.edu)
------------------------------
Date: Thu, 2 Aug 90 08:14:39 CDT
From: cpotter@ncsa.uiuc.edu (Clint Potter)
Subject: Re: Displaying 24bit/pix images on 8bit/pix device
This is in regards to the Vision-List request for packing 24bit RGB images.
The National Center for Supercomputing Applications has developed a utility
for this called r24r8. Source code is available via anonymous
ftp at ftp.ncsa.uiuc.edu (128.174.20.50) from the directory
unsupported/Icpack.
Clint Potter
National Center for Supercomputing Applications
Universtity of Illinois at Urbana-Champaign
r24r8 Quantizes a raw RGB 24 bit "pixel" image into a 8 bit image
with RGB palette.
usage: r24r8 x_dim y_dim r24_file r8_file pal_file
On Input:
--------
x_dim - X dimension of image (horizontal size).
y_dim - Y dimension of image (vertical size).
r24_file - File containing the raw RGB 24
bit image. The order of pixels is left to
right and top to bottom. Each pixel is
three contiguous bytes: red byte, green byte,
and blue byte. Use filter ptox to convert
from "planar" to "pixel" where planar means
all red bytes for the whole image, followed
by all green bytes for the whole image,
followed by all blue bytes for the whole
image.
On Output:
---------
r8_file - File where 8 bit image is written.
pal_file - File where 768 (256*3) byte palette
is written. Order is 256 bytes of red,
256 of green, and 256 bytes of blue.
by: NCSA
date(s): May 89, Jun 89, Aug 89
------------------------------
Date: Thu, 2 Aug 90 14:53:27 BST
From: cs_s424@ux.kingston.ac.uk
Subject: Re: Converting 24 bit images to 8 bit.
Organisation: Kingston Polytechnic
> We have a few images in 24 bits/pixel format (one byte each of Red, Green
> and Blue per pixel). We need to convert these images into sun raster 8 bit
> color format or GIF format in order to display on X windows. Is there any
> available programs or formulae that would let us either convert these images
> to a compatible form or let us display them directly on the X windows?
> Any information in this regard will be greatly appreciated.
I wrote a couple of programs to display 24 bit images on an 8 bit device. The
first program used a very crude method but worked adequately. Basically, you
need to scale down 24 bits into 8. The way to do this is like so:
blue green red
|_______|___________|___________|
| 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
The red in the first 3 bits, then the green in second 3 bits and
finally the blue (which the eye is least sensitive to) in the last 2.
When we read the 24 bit data, we have to scale it down into these
bits. Then we have to set up the palette or Colour Look-Up Table
(CLUT) on the 8 bit device to display the right colour when presented
with this bit pattern. Once the pallete is set, we can read in the
image data, put it in the above format and then display the pixel.
This is a simple method, but alot of information is lost. I then wrote
a second program to set up the CLUT to represent what colours are in
the image, rather than doing a conversion, as above. In my case, I had
three 8 bit images, one containing the red, the blue and green
components. (This was because I was taking colour images using an
8-bit frame grabber but placing a red filter in front of the lense,
grabbing, then a green filter, grab, then lastly the blue filter, and
grab, leaving me with the RGB components of a colour image in three
separate images.)
What the program does is read in all the red, green and blue
components, and puts them into a 'colour cube'. This is scanned for
the most commonly used colour groups and the palette is set up based
on this information. Then the image is scanned and a reference file is
set up to index the palette. This gives you two files; the palette and
the image where each pixel references the palette The program was
written on a Sun.
This program produced incredible results. Later, I was able to display
the image on a true 24 bit display, and you could see little
difference. The method will begin to fall down if the image is
especially rich in different colours. [Code is available,]
Unfortunately, it is a little hacked together, you will have to make
some modifcations on how the image is read. Hope this is of help.
[ Code can be obtained directly from P.J.Netherwood@kingston.ac.uk.
phil... ]
Paul Netherwood janet : P.J.Netherwood@uk.ac.kingston
Research internet : P.J.Netherwood@kingston.ac.uk
phone : (+44) 81 549 1366 ext 2923
local : cs_s424@ux.king
School of Computer Science and Electronic Systems,
Kingston Polytechnic, Penrhyn Road, Kingston-upon-Thames, Surrey KT1 2EE, UK.
------------------------------
Date: Thu, 2 Aug 90 08:26:29 BST
From: Adrian F Clark <alien%sx.ac.uk@pucc.PRINCETON.EDU>
Subject: Re: Displaying 24bit/pix images on 8bit/pix device
In Vision-List of Wed Aug 01, Surajit Chekavarti writes:
> We have a few images in 24 bits/pixel format (one byte each of Red, Green
> and Blue per pixel). We need to convert these images into sun raster 8 bit
> color format or GIF format in order to display on X windows. Is there any
> available programs or formulae that would let us either convert these images
> to a compatible form or let us display them directly on the X windows?
>
> Any information in this regard will be greatly appreciated.
There is a widely-used algorithm for doing this due to Paul Heckbert:
"Image Quantization for Frame Buffer Display" in the SIGGRAPH '82
proceedings (p297). For code to implement it, look in Jef Poskanzer's
PBMPLUS package, available from your nearest X11R4 archive site. The
most recent distributed version is dated 22 Nov 1989.
Adrian F. Clark
JANET: alien@uk.ac.essex ARPA: alien%uk.ac.essex@nsfnet-relay.ac.uk
BITNET: alien%uk.ac.essex@ac.uk PHONE: (+44) 206-872432 (direct)
Dept ESE, University of Essex, Wivenhoe Park, Colchester, Essex, C04 3SQ, UK.
------------------------------
Date: Sun, 5 Aug 90 10:21:07 EDT
From: black@seismo.CSS.GOV (Mike Black)
Subject: Re: Displaying 24bit/pix images on 8bit/pix device
Organization: Center for Seismic Studies, Arlington, VA
I think you can get fbmplus.tar.Z from ucsd.edu. It does many manipulations
including converting to/from Sun raster files and modifying color maps.
(I don't seem to have ftp available or I would've confirmed this)
Mike...
: usenet: black@beno.CSS.GOV : land line: 407-494-5853 : I want a computer:
: real home: Melbourne, FL : home line: 407-242-8619 : that does it all!:
------------------------------
Date: Sat, 4 Aug 90 17:49:19 bst
From: Eduardo Bayro <eba@computing-maths.cardiff.ac.uk>
Subject: feature extraction
Organization: Univ. of Wales Coll. of Cardiff, Dept. of Electronic & Systems Engineering
Dear frends!
I am a research fellow involved with application of neural networks
for low level processing. Neural Networks require a short feature
vector. I appreciate if anybody can suggest me ways to extract
features of 2D . I am trying to recognize shape s and be abble to
identify faulty objects. The procedure to prepare the vector can be
based in classical algorithms provided that are not time consuming. I
am working with microsoft C 5.1 ( full IBM/AT compatible
microcomputer. I will be thank full if anybody send me source code for
a feature extraction procedure and its documentation. Any results I
will let know many thanks, Eduardo.
Eduardo Bayro, School of Electrical, Electronic and Systems Engineering,
University of Wales College of Cardiff, Cardiff, Wales, UK.
Internet: eba%cm.cf.ac.uk@nsfnet-relay.ac.uk Janet: eba@uk.ac.cf.cm
UUCP: eba@cf-cm.UUCP or ...!mcsun!ukc!cf-cm!eba
------------------------------
Date: 6 Aug 90 13:40 -0700
From: Esfandiar Bandari <bandari@cs.ubc.ca>
Subject: landscaping data
Hello:
I am looking for a program that generates three dimensional
landscape dislays -- i.e. given a x-y grid of points (or preferably a
number of random [x,y] coordinates points) and thier respective Z (or
distance form the plain) values, the program would then generate a 3D
wiremesh (and solid surfaces) representation with hidden surface
removal.
There is a program called SPOOF that does this but I am not
sure if it works with X windows and how to get the source to it. We
also have two programs here that work for Iris's but again nothing for
the X (and they are not callable from C).
Any pointers to the electronic source(s) would be appreciated.
Thanks in advance.
--- Esfandiar
------------------------------
Date: Thu, 2 Aug 90 15:32:57 PDT
From: shapiro@lillith.ee.washington.edu (Linda Shapiro)
Subject: change in CVGIP Journal
ANNOUNCMENT
Computer Vision, Graphics, and Image Processing has grown to cover the work
of two somehwat diverse communities -- computer graphics and image processing
and image understanding. In order to serve the needs of these distinct
disciplines, the journal will be divided into two publications in 1991:
CVGIP: Graphical Models and Image Processing
CVGIP: Image Understanding
Each journal will feature original research papers. Expository and review
papers and papers that embody novel concepts in applications and techniques
will also be considered.
CVGIP: Image Understanding
Linda G. Shapiro, Editor-in-Chief
Area Editors and Area
Ruzena K. Bajcsy, Shape
Larry S. Davis, Matching and Recognition
Herbert Freeman, Special Topics
Robert M. Haralick, Theory
Thomas S. Huang, Motion
Ramesh Jain, Range
Edward Riseman, Vision Systems
Azriel Rosenfeld, Early Vision
Hanan Samet, Data Structures
Steven L. Tanimoto, Architecture and Language
FOCUS
The central focus of this journal is the computer analysis of pictorial
information. CVGIP: Image Understanding will publish papers covering all
aspects of image analysis from the low-level, iconic processes of early
vision to the high-level symbolic processes of recognition and interpretation.
AIM
CVGIP: Image Understanding will distinguish itself from other journals in
the computer vision field by covering a wide range of topics in the image
understanding area, including insights that differ from predominant views.
Letters to the Editor expressing stimulating viewpoints and philosophies
will also be published. Suitability for publication will be based on
relevance to the overall IU editorial goals, originality, technical quality,
importance, and readability.
MANUSCRIPTS for CVGIP: Image Understanding should be submitted in
quadruplicate to:
CVGIP: Image Understanding
Editorial Office
1250 Sixth Avenue
San Diego, CA 92101
CVGIP: Graphical Models and Image Processing
Norman Badler, Editor-in-Chief
Rama Chellappa, Editor-in-Chief
Area Editors
Brian A. Barsky
Gabor Herman
R.L. Kashyap
Arun N. Netravali
Franco P. Preparata
Aristides A. Requicha
Demetri Terzopoulos
John W. Woods
FOCUS
The synthesis methods and computational models underlying computer-generated
or -processed imagery are the central focus of CVGIP: Graphical Models and
Image Processing.
RESEARCH AREAS
Image Synthesis
Visualization and physical models
Curves and surfaces
Solid models
Animation
Computational geometry
Three-dimensional imaging
Image models and transforms
Enhancement and restoration
Compression
AIM
The Editors-in-Chief, Norman Badler and Rama Chellappa, will assume editorial
responsibility for the areas of graphical models and image processing,
respectively. Area editors, selected from among the premier researchers in their
fields, will oversee the editorial contents of each of the above topics.
This specialization will ensure prompt review of submitted manuscripts.
Suitability for publication will be based on relevance to the overall GMIP
editorial goals, originality, technical quality, importance, and readability.
MANUSCRIPTS for CVGIP: Graphical Models and Image Processing should be
submitted in quadruplicate to
CVGIP: Graphical Models and Image Processing
Editorial Office
1250 Sixth Avenue
San Diego, CA 92101
------------------------------
End of VISION-LIST
********************
