 |
|
| Science Forum Index » Image Processing Forum » Latest Version of Tria Image Processing (It's a good... |
|
Page 1 of 1 |
|
| Author |
Message |
| JimAtQuarktet... |
 Posted: Fri Oct 23, 2009 7:34 am |
|
|
|
Guest
|
Hi All,
We are very proud to announce the release of Tria, version 2.6. It
contains all the functions and tools that we want from an image
processing program. This latest version has improved SeDDaRA blind
deconvolution processing, fast image registration (image alignment)
for translations, rotations, and scale, mouse tools that allow the
user to measure distances and angles between points and determine the
size and location of spots, a nifty image magnifier,region-of-interest
stats and contrast, batch processing of all major functions, and batch
averaging. See http://www.quarktet.com/Tria.html for more information. |
|
|
| Back to top |
|
|
|
| aruzinsky... |
| Posted: Sat Oct 24, 2009 6:46 am |
|
|
|
Guest
|
On Oct 23, 11:34 am, JimAtQuarktet <ca... at (no spam) quarktet.com> wrote:
[quote]Hi All,
We are very proud to announce the release of Tria, version 2.6. It
contains all the functions and tools that we want from an image
processing program. This latest version has improved SeDDaRA blind
deconvolution processing, fast image registration (image alignment)
for translations, rotations, and scale, mouse tools that allow the
user to measure distances and angles between points and determine the
size and location of spots, a nifty image magnifier,region-of-interest
stats and contrast, batch processing of all major functions, and batch
averaging. Seehttp://www.quarktet.com/Tria.htmlfor more information.
[/quote]
Looking at your deconvolution examples, I came across
http://www.quarktet.com/Dx-group.jpg, which is awful, even by layman
standards. The flash bulb reflections on people's eyeglasses became
bigger after processing. Duh? Using my method, I deconvolved your
"Before" image, http://www.quarktet.com/group.jpg, and got
http://www.general-cathexis.com/images/arock-blurredPseudoLAD.jpg.
More generally, were your actual input images the JPEGs displayed in
your gallery or versions with less lossy compression and maybe higher
resolutions? |
|
|
| Back to top |
|
|
|
| JimAtQuarktet... |
| Posted: Sun Oct 25, 2009 5:31 pm |
|
|
|
Guest
|
[quote]Looking at your deconvolution examples, I came acrosshttp://www.quarktet.com/Dx-group.jpg, which is awful, even by layman
standards. The flash bulb reflections on people's eyeglasses became
bigger after processing. Using my method, I deconvolved your
"Before" image, http://www.quarktet.com/group.jpg, and gothttp://www.general-cathexis.com/images/arock-blurredPseudoLAD.jpg.
More generally, were your actual input images the JPEGs displayed in
your gallery or versions with less lossy compression and maybe higher
resolutions?
[/quote]
I agree that the group shot needs some work. In addition to the
reflections off the eyeglasses, there is a glow around the heads of
the people in the back row. This is clear indication that the PSF
used in the deconvolution was less than ideal. I will have to revisit
that image.
Concerning the images displayed in the gallery, some are the actual
images that were processed, and others have been downsized and
compressed. The group image, as I recall, is the actual image, and I
think that holds for all the pictures on that page. The NASA
Wilkinson Microwave Anistropy Probe image was a 2k by 2k grayscale 8-
bit image, which had to be condense to fit on the page. The change in
the false color representation is mostly due to the improved contrast
in the image. The X-rays were taken off the web as jpegs and are
displayed in the resolution that I processed. The image of fruit on
the misc. page benefited from being a 16-bit image, which can not be
displayed. When the images are saved as jpegs, I do use the the least
amount of compression possible.
In general, I do try to preserve as much of the original image as
possible when I post. If my methods are as good as I have been
saying, then I want others to download and process the gallery images
for comparison with their own methods. It is easy enough to add an
indicator when the originals are displayed. |
|
|
| Back to top |
|
|
|
| aruzinsky... |
| Posted: Mon Oct 26, 2009 4:51 am |
|
|
|
Guest
|
JimAtQuarktet wrote:
[quote]Looking at your deconvolution examples, I came acrosshttp://www.quarktet.com/Dx-group.jpg, which is awful, even by layman
standards. The flash bulb reflections on people's eyeglasses became
bigger after processing. Using my method, I deconvolved your
"Before" image, http://www.quarktet.com/group.jpg, and gothttp://www..general-cathexis.com/images/arock-blurredPseudoLAD.jpg.
More generally, were your actual input images the JPEGs displayed in
your gallery or versions with less lossy compression and maybe higher
resolutions?
I agree that the group shot needs some work. In addition to the
reflections off the eyeglasses, there is a glow around the heads of
the people in the back row. This is clear indication that the PSF
used in the deconvolution was less than ideal. I will have to revisit
that image.
Concerning the images displayed in the gallery, some are the actual
images that were processed, and others have been downsized and
compressed. The group image, as I recall, is the actual image, and I
think that holds for all the pictures on that page. The NASA
Wilkinson Microwave Anistropy Probe image was a 2k by 2k grayscale 8-
bit image, which had to be condense to fit on the page. The change in
the false color representation is mostly due to the improved contrast
in the image. The X-rays were taken off the web as jpegs and are
displayed in the resolution that I processed. The image of fruit on
the misc. page benefited from being a 16-bit image, which can not be
displayed. When the images are saved as jpegs, I do use the the least
amount of compression possible.
In general, I do try to preserve as much of the original image as
possible when I post. If my methods are as good as I have been
saying, then I want others to download and process the gallery images
for comparison with their own methods. It is easy enough to add an
indicator when the originals are displayed.
[/quote]
Surely, you know that almost all web browsers can display 16 bit/
channel PNG which only use lossless compression. If you want others to
download and process your input images, you shouldn't display them so
that mouse cursor passover changes the display to the output image
because that prevents right clicking the input image and using "save
as" from the popup menu. You can write Java script to change the
display by left clicking a button instead of cursor passover of
image. I had to look at your page source code to get your input
image URL. |
|
|
| Back to top |
|
|
|
|
|
All times are GMT - 5 Hours
The time now is Wed Dec 09, 2009 4:55 pm
|
|