Feature #767

Support viewing of images containing data (not neccessary RGB) values.

Added by Christian over 2 years ago. Updated over 2 years ago.

Status:NewStart date:07.12.2016
Priority:NormalDue date:
Assignee:-% Done:

0%

Category:-
Target version:-
Version: Operating System:

Description

I often compute image maps that contain negative values, e.g. to indicate missing values, while still having the whole 8bit range available for data values.

In order to inspect such images, I currently rely on plotting tools of matplotlib, or matlab,
as no image viewer I know of is capable of handling such images gracefully.

- Scale the whole image data range to visble range.
- Indicate to user that the image is shown in pseudo-color.
- In Pseudocolor toolbar allow images to be shown using some colormap.

History

#1 Updated by Markus over 2 years ago

I am not sure, if I do understand fully.
How do you save negative values to 8-bit images?
We provide a nomacs matlab script: http://nomacs.org/download/ maybe this is of interest to you.
The script allows you to visualize data from matlab using nomacs

>> data = rand(10, 10)-0.5;
>> nomacs(data);    % just show the data (negative values are clipped)
>> nomacs(data,[]); % scale the image to the visible range

#2 Updated by Christian over 2 years ago

How do you save negative values to 8-bit images?

I don't.

Naturally for differential data here is always a transfer function f(I):[-x,x] -> [0,255], that has to be applied prior saving, and after loading.

The still case stands for formats that support negative values, e.g. EXR, PFM, though. Also @

We provide a nomacs matlab script: http://nomacs.org/download/ maybe this is of interest to you.
The script allows you to visualize data from matlab using nomacs.

Basically I asked to have what the script does to have within nomacs.

I find the plotting tools that available under matlab, or matplotlib in the python case, lacking for interactive use.
So instead of writing the $x+1$th visualize-and-save-for-later wrapper around $imwrite$, I save stuff as raw as possible as image.

Gimps Curves tool can do wonders in getting insights over arbritrary image-like data. Moreover you have 4 potentially independent channels!

Currently I use nomacs as default app for all day-to-day image file handling. Together with an informative histogram, and the colormap function nomacs is just one step away from beeing my visual debugger.

Also available in: Atom PDF