Fwd: Best path to take

Posted by Robert Lockwood on Oct 10, 2013; 8:28pm
URL: http://imagej.273.s1.nabble.com/Best-path-to-take-tp5005126p5005127.html

>> Why not just save the images in TIFF format? Why invent another file
format?


When I saw that I could save the files as TIFF I gave it a try but that
destroys the original data. What's is apparently saved is the image data,
that is, it's reduced to  BufferedImage.  That was a disappointing
exercise.

I know that it is possible for a TIFF file to have the original data in it
but AFAIK Fiji doesn't support writing this Scientific data mode, alas.

---------- Forwarded message ----------
From: Rasband, Wayne (NIH/NIMH) [E] <[hidden email]>
Date: Thu, Oct 10, 2013 at 1:08 PM
Subject: Re: Best path to take
To: Robert Lockwood <[hidden email]>


Dear Robert,

Why not just save the images in TIFF format? Why invent another file format?

Best regards,

-wayne


On Oct 10, 2013, at 3:29 PM, Robert Lockwood wrote:

> Johannes, thanks for your help with my initial attempt to write a plugin.
>
> Before I proceed I am asking you and others in the group for some advice.
>
> My situation is that we are developing a camera system for in house
> infrared photography, primarily of wild fires.  At this point I need to be
> able to provide a way for others to view our image data and thought that
> Fiji would be a good solution.
>
> I've decided that our images will all have a metadata header and the first
> 24 bytes would be identical in function.  Our image data is all (C-talk
> here) unsigned shorts with the 12-14 bits of precision depending on the
> camera and are monochrome.
>
> The size varies greatly among the cameras with a probable minimum of 512
> lines by 640 pixels up to 11 megapixels.  The image data for most of the
> cameras are very low contrast unless fire is in the image and consist of
> raw DN values.
>
> The common part of the header, the minimum information for the image to be
> displayed, is"  Offset-to-image-data, number-of-lines,
> number-of-samples-per-line, pixel-size, and time-of-acquisition (Java date
> time in ms).  The first four data are Java ints, and the time is a Java
> long for a total of 24 bytes.
>
> I'd like to tell the few people who need to view the images to just
> download and install Fiji on their LINUX, Macs, and Windows machines, and
> then how to load and use the plugin/macro I supply.  The object is for the
> user to use Fiji to browse and display  our  files (possible extension
> .rawimage) and not need to enter any image metadata as they may not know
> those data.  Once the image is in Fiji it can be manipulated, saved-as,
> etc.  In essence it should work analogous to importing a raw image without
> needing to input metadata by hand.
>
> TIA for any insight, pointers, etc.
>
> Nate
>
> --
> When I was 12 I thought I would live forever.
> So far, so good.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
When I was 12 I thought I would live forever.
So far, so good.

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html