3D z-scale in heigh maps

[fisavi]

Hello,

I will be very grateful if someone can help me.

I'm using the "interactive 3D surface plot" plugin to do a 3D  
representation
of a surface map (not stacks, only a grey-scaled image). The problem is
cannot change the Z-axis scale in the representation using
"Analyze->Calibrate" or any other thing I have tried.

The thing is that, when I open the image in the .nrrd format, it's scaled,
but from -0.3e-5 to 0.85e-6 meters (I need it in microns). If I convert it
to 16 bits, it always goes from 0 to 65536, obviously.

Thanks

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

[Michael Schmid]

Hi Fidel,

you can convert your image to Float (32 bit). The float values will be equal to the calibrated 8-bit or 16-bit values of the original image and shown as such in the interactive 3D surface plot.

Michael
________________________________________________________________
On Jun 18, 2013, at 15:12, Fidel Salas wrote:

> Hello,
>
> I will be very grateful if someone can help me.
>
> I'm using the "interactive 3D surface plot" plugin to do a 3D representation
> of a surface map (not stacks, only a grey-scaled image). The problem is
> cannot change the Z-axis scale in the representation using
> "Analyze->Calibrate" or any other thing I have tried.
>
> The thing is that, when I open the image in the .nrrd format, it's scaled,
> but from -0.3e-5 to 0.85e-6 meters (I need it in microns). If I convert it
> to 16 bits, it always goes from 0 to 65536, obviously.
>
> Thanks

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

[fisavi]

Thanks. It works.

My original image was in 32 bits (sorry, I didn't say that in my previous post). I converted it to 16 bits and applied a new calibration. After that, I reconverted it to 32 bits and...voila!!, now I have the correct scale in the z-axis.

But, is there a way to change the scale in the original 32-bits image directly?.

Fidel

[Michael Schmid]

Hi Fidel,

ImageJ does not allow calibration of float (32-bit) images with Analyze>Calibrate.
In other words, for float images, the pixel values and their meaning in physical units cannot be different.

If you want to convert float images to a (different) calibration, you have to manipulate the pixel data, e.g. Process>Math>multiply by 1000 to convert from meters to millimeters.

Michael
________________________________________________________________
On Jun 18, 2013, at 18:33, fisavi wrote:

> Thanks. It works.
>
> My original image was in 32 bits (sorry, I didn't say that in my previous
> post). I converted it to 16 bits and applied a new calibration. After that,
> I reconverted it to 32 bits and...voila!!, now I have the correct scale in
> the z-axis.
>
> But, is there a way to change the scale in the original 32-bits image
> directly?.
>
> Fidel

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

[fisavi]

Hi Michael,

I had tried that, but the image was "visually" converted to a black-and-white image or to an all-gray one (the effect will depend on the mathematical operation), and the 3d representation didn't work. Of course, the pixel values were the good ones. More curious: if I open the, e.g., "brightness/contrast" or "threshold" window and move the sliders but without applying the changes, the image recovers its aspect and the 3d representation works properly.

There seems to be necessary some kind of "actualize changes".

Many thanks,
Fidel

[Michael Schmid]

Hi Fidel,

if the image does not display as it should do, simply press 'Reset' or 'Auto' on the 'Brightness&Contrast' panel. For float (32-bit) images, obviously there is nothing like a 'natural' range of pixel values, such as 0-255 for 8-bit data.

It seems to me that the interactive 3D surface plot only uses the image data between the upper and lower bounds of 'Brightness&Contrast'.

Michael
________________________________________________________________
On Jun 19, 2013, at 14:46, fisavi wrote:

> Hi Michael,
>
> I had proved that, but the image was "visually" converted to a
> black-and-white image or to an all-gray one (the effect will depend on the
> mathematical operation), and the 3d representation didn't work. Of course,
> the pixel values were the good ones. More curious: if I open the, e.g.,
> "brightness/contrast" or "threshold" window and move the sliders but without
> applying the changes, the image recovers its aspect and the 3d
> representation works properly.
>
> There seems to be necessary some kind of "actualize changes".
>
> Many thanks,
> Fidel

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