Thread Subject: How to create Fluorescence images

Hello

Gray 2 RGB ...fine
RGB 2 gray fine

But how can one create Fluorescence images in MATLAB

Input = Gray Scale Image
Output = Fluorescence image

Any tutorials would help me learn

Kindly advise
Ash

Ash wrote:
>
>
> Hello
>
> Gray 2 RGB ...fine
> RGB 2 gray fine
>
> But how can one create Fluorescence images in MATLAB
>
> Input = Gray Scale Image
> Output = Fluorescence image
>
> Any tutorials would help me learn
>
> Kindly advise
> Ash
  
shine a uv light on your monitor?

dave wrote:
>
>
> Ash wrote:
>>
>>
>> Hello
>>
>> Gray 2 RGB ...fine
>> RGB 2 gray fine
>>
>> But how can one create Fluorescence images in MATLAB
>>
>> Input = Gray Scale Image
>> Output = Fluorescence image
>>
>> Any tutorials would help me learn
>>
>> Kindly advise
>> Ash
>
> shine a uv light on your monitor?

Nah. Plug your monitor into a 440 volt,
3 phase outlet. (If you can find one in
the vicinity.) It will probably shine
quite brightly over all wavelengths,
even though only for a very short time.

For the OP, please define what you mean
by a "fluorescence image".

John

Thanks John and Dave

the images must i get are the ones are the likes in

 <http://www.ii.bham.ac.uk/clinicalimmunology/CISimagelibrary/>

Thanks
Ash

John D'Errico wrote:
>
>
> dave wrote:
>>
>>
>> Ash wrote:
>>>
>>>
>>> Hello
>>>
>>> Gray 2 RGB ...fine
>>> RGB 2 gray fine
>>>
>>> But how can one create Fluorescence images in MATLAB
>>>
>>> Input = Gray Scale Image
>>> Output = Fluorescence image
>>>
>>> Any tutorials would help me learn
>>>
>>> Kindly advise
>>> Ash
>>
>> shine a uv light on your monitor?
>
> Nah. Plug your monitor into a 440 volt,
> 3 phase outlet. (If you can find one in
> the vicinity.) It will probably shine
> quite brightly over all wavelengths,
> even though only for a very short time.
>
> For the OP, please define what you mean
> by a "fluorescence image".
>
> John

Ash wrote:
>
>
> Thanks John and Dave
>
> the images must i get are the ones are the likes in
>
> <http://www.ii.bham.ac.uk/clinicalimmunology/CISimagelibrary/>
>
> Thanks
> Ash
>
> John D'Errico wrote:
>>
>>
>> dave wrote:
>>>
>>>
>>> Ash wrote:
>>>>
>>>>
>>>> Hello
>>>>
>>>> Gray 2 RGB ...fine
>>>> RGB 2 gray fine
>>>>
>>>> But how can one create Fluorescence images in MATLAB
>>>>
>>>> Input = Gray Scale Image
>>>> Output = Fluorescence image
>>>>
>>>> Any tutorials would help me learn
>>>>
>>>> Kindly advise
>>>> Ash
>>>
>>> shine a uv light on your monitor?
>>
>> Nah. Plug your monitor into a 440 volt,
>> 3 phase outlet. (If you can find one in
>> the vicinity.) It will probably shine
>> quite brightly over all wavelengths,
>> even though only for a very short time.
>>
>> For the OP, please define what you mean
>> by a "fluorescence image".
>>
>> John
  
well, normally to get those you use a microscope with an uv light
source and take a picture of the results. flourescence is not
something that can be simulated by just processing an image, it is a
property of the material being exposed to a light source. i.e. a
'white' material viewed under room lighting may glow green, red,
blue, or just about any other color under uv light depending on what
it is made from.

dave wrote:
>
> well, normally to get those you use a microscope with an uv light
> source and take a picture of the results. flourescence is not
> something that can be simulated by just processing an image, it is
> a
> property of the material being exposed to a light source. i.e. a
> 'white' material viewed under room lighting may glow green, red,
> blue, or just about any other color under uv light depending on
> what
> it is made from.
  
Yes, I understand what fluorecence
means. My point was the OP apparently
wants something that is not simply
obtainable from a gray scale image.
So we need to understand their goals.

 <http://en.wikipedia.org/wiki/Fluorescence>

John

Thanks Dave,
that helped. But then the image link i forward has images which are
in jpg/bmp etc etc format and they are images which were captured
from a source like you mentioned.

But they have a colormap which would be quite specific/characteristic
to fluorescences images. Cant we simulate that?

But i guess what you are saying is right. i will check with the
source hardware description.
Thank you
Ash

dave wrote:
>
>
> Ash wrote:
>>
>>
>> Thanks John and Dave
>>
>> the images must i get are the ones are the likes in
>>
>> <http://www.ii.bham.ac.uk/clinicalimmunology/CISimagelibrary/>
>>
>> Thanks
>> Ash
>>
>> John D'Errico wrote:
>>>
>>>
>>> dave wrote:
>>>>
>>>>
>>>> Ash wrote:
>>>>>
>>>>>
>>>>> Hello
>>>>>
>>>>> Gray 2 RGB ...fine
>>>>> RGB 2 gray fine
>>>>>
>>>>> But how can one create Fluorescence images in
MATLAB
>>>>>
>>>>> Input = Gray Scale Image
>>>>> Output = Fluorescence image
>>>>>
>>>>> Any tutorials would help me learn
>>>>>
>>>>> Kindly advise
>>>>> Ash
>>>>
>>>> shine a uv light on your monitor?
>>>
>>> Nah. Plug your monitor into a 440 volt,
>>> 3 phase outlet. (If you can find one in
>>> the vicinity.) It will probably shine
>>> quite brightly over all wavelengths,
>>> even though only for a very short time.
>>>
>>> For the OP, please define what you mean
>>> by a "fluorescence image".
>>>
>>> John
>
> well, normally to get those you use a microscope with an uv light
> source and take a picture of the results. flourescence is not
> something that can be simulated by just processing an image, it is
> a
> property of the material being exposed to a light source. i.e. a
> 'white' material viewed under room lighting may glow green, red,
> blue, or just about any other color under uv light depending on
> what
> it is made from.

Ash wrote:
>
>
> Thanks Dave,
> that helped. But then the image link i forward has images which are
> in jpg/bmp etc etc format and they are images which were captured
> from a source like you mentioned.
>
> But they have a colormap which would be quite
> specific/characteristic
> to fluorescences images. Cant we simulate that?
>
> But i guess what you are saying is right. i will check with the
> source hardware description.
> Thank you
> Ash

Then all you want is a black to green
colormap?

t = linspace(0,1,256)';
cm = t*[0 1 0];

[x,y] = meshgrid(0:100);
surf(x.^2 + y.^2);
colormap(cm)
shading interp

John

Thats pretty much it!!!
I shall apply and tell you what i get as output
Thanks John
Ash
 
John D'Errico wrote:
>
>
> Ash wrote:
>>
>>
>> Thanks Dave,
>> that helped. But then the image link i forward has images which
> are
>> in jpg/bmp etc etc format and they are images which were
> captured
>> from a source like you mentioned.
>>
>> But they have a colormap which would be quite
>> specific/characteristic
>> to fluorescences images. Cant we simulate that?
>>
>> But i guess what you are saying is right. i will check with the
>> source hardware description.
>> Thank you
>> Ash
>
> Then all you want is a black to green
> colormap?
>
> t = linspace(0,1,256)';
> cm = t*[0 1 0];
>
> [x,y] = meshgrid(0:100);
> surf(x.^2 + y.^2);
> colormap(cm)
> shading interp
>
> John

Maybe also see

 <http://blogs.mathworks.com/steve/?p=57>
and
 <http://bsp.pdx.edu/>

which discuss conversion between RGB colour and grey scale while
preserving luminance.

Thanks Malcolm...those sites are informative.
Shall write about my outcome
 Malcolm Lidierth wrote:
>
>
> Maybe also see
>
> <http://blogs.mathworks.com/steve/?p=57>
> and
> <http://bsp.pdx.edu/>
>
> which discuss conversion between RGB colour and grey scale while
> preserving luminance.

Ash,

often people working in fluorescence microscopy want to visualise
more than one fluorescent probe. Typically this is done by combining
multiple gray-scale images in a false colour image. Is this what you
are looking for?

In principle, this is a trivial operation (it can simply be done
using cat) but can get more complicated if you need to rescale the
separate images before combining them.

Cheers,

Alex

The views expressed in this e-mail are those of the originator and do
not necessarily represent the views of NPL Management Ltd. Nothing in
this e-mail shall bind NPL Management Ltd in any contract or
obligation.

Thank you Alex.
Actually i am analysing microarray data. But when in gray scale some
of the important features are not very clear which i think will
become clear when i move it to a fluorescence image. Why
fluorescence? cause gray to rgb doe s not seem to highlight that.
like the examples i saw on the page i linked above.
More than one fluorescent probe ? would that be a Red/Green/Blue
mapping then?
I did go through the tutorials above and they helped.
kindly let me know if you have any tutorials on your suggestion.

Thank you
Ash
 
Alex Knight wrote:
>
>
> Ash,
>
> often people working in fluorescence microscopy want to visualise
> more than one fluorescent probe. Typically this is done by
> combining
> multiple gray-scale images in a false colour image. Is this what
> you
> are looking for?
>
> In principle, this is a trivial operation (it can simply be done
> using cat) but can get more complicated if you need to rescale the
> separate images before combining them.
>
> Cheers,
>
> Alex
>
> The views expressed in this e-mail are those of the originator and
> do
> not necessarily represent the views of NPL Management Ltd. Nothing
> in
> this e-mail shall bind NPL Management Ltd in any contract or
> obligation.

Hi Ash,

Microarrays are a good example of what I meant, as people are usually
interested in measuring ratios. So typically you take two images of
"red" and "green" fluorescence and combine them as a false colour
image. So in this example you would take two grayscale intensity
images using two different sets of filters and lasers and combine
them as an RGB image.

In MATLAB this would be something like

rgbim = cat( 3, redim, greenim, blueim )
where blueim would be something like:
zeros( size( redim ) )
as there is no blue channel in this instance.

Note that often the "green" image is Cy3 (which is actually orange
fluorescence) and the "red" is Cy5 (which is indeed red fluorescence)
and you will see that the whole colour thing is somewhat arbitrary.

Now if you have a single grayscale image and you map it to the green
layer of an RGB image, it should not make too much difference to the
level of detail. this is more down to what you do to the contrast,
e.g. with imadjust.

Have you thought about using a colourmap like jet? This can be very
useful for seeing details that are not obvious in a grayscale image.

Cheers,

Alex

The views expressed in this post are those of the originator and do
not necessarily represent the views of NPL Management Ltd. Nothing in
this post shall bind NPL Management Ltd in any contract or
obligation.

Ash wrote:
>
>
> Thank you Alex.
> Actually i am analysing microarray data. But when in gray scale
> some
> of the important features are not very clear which i think will
> become clear when i move it to a fluorescence image. Why
> fluorescence? cause gray to rgb doe s not seem to highlight that.
> like the examples i saw on the page i linked above.
> More than one fluorescent probe ? would that be a Red/Green/Blue
> mapping then?
> I did go through the tutorials above and they helped.
> kindly let me know if you have any tutorials on your suggestion.
>
> Thank you
> Ash
>

Thank you Dr.Knight,
Your explanation and suggestion was helpful. I shall put my images
through some related algorithms.
Regards
Ash

Alex Knight wrote:
>
>
> Hi Ash,
>
> Microarrays are a good example of what I meant, as people are
> usually
> interested in measuring ratios. So typically you take two images of
> "red" and "green" fluorescence and combine them as a false colour
> image. So in this example you would take two grayscale intensity
> images using two different sets of filters and lasers and combine
> them as an RGB image.
>
> In MATLAB this would be something like
>
> rgbim = cat( 3, redim, greenim, blueim )
> where blueim would be something like:
> zeros( size( redim ) )
> as there is no blue channel in this instance.
>
> Note that often the "green" image is Cy3 (which is actually orange
> fluorescence) and the "red" is Cy5 (which is indeed red
> fluorescence)
> and you will see that the whole colour thing is somewhat arbitrary.
>
> Now if you have a single grayscale image and you map it to the
> green
> layer of an RGB image, it should not make too much difference to
> the
> level of detail. this is more down to what you do to the contrast,
> e.g. with imadjust.
>
> Have you thought about using a colourmap like jet? This can be very
> useful for seeing details that are not obvious in a grayscale
> image.
>
> Cheers,
>
> Alex
>
> The views expressed in this post are those of the originator and do
> not necessarily represent the views of NPL Management Ltd. Nothing
> in
> this post shall bind NPL Management Ltd in any contract or
> obligation.
>
> Ash wrote:
>>
>>
>> Thank you Alex.
>> Actually i am analysing microarray data. But when in gray scale
>> some
>> of the important features are not very clear which i think will
>> become clear when i move it to a fluorescence image. Why
>> fluorescence? cause gray to rgb doe s not seem to highlight
that.
>> like the examples i saw on the page i linked above.
>> More than one fluorescent probe ? would that be a
Red/Green/Blue
>> mapping then?
>> I did go through the tutorials above and they helped.
>> kindly let me know if you have any tutorials on your
suggestion.
>>
>> Thank you
>> Ash
>>

This is a thread that was discussed a whie back, but I
have a related question so thought I'd use it.

What I need to do is similar to what has been discussed
here, with a difference.

I have: two grayscale images from a microscope. One of
them is the background and should stay in gray, and the
other is a fluorescent image of some of the objects.

The fluorescent image is taken with a specific filter, and
I would like to pseudo color it (according to the specific
filter) and overlay on top of the gray image.

How can I overlay those two and have the fluorescent part
nice and bright?
I have tried subimage and having two colormaps, but this
gives a dark image.
We are able to do this nicely with Photoshop, and would
like to do it in near real-time with the microscope, using
matlab.

Any help would be appreciated..

Daphne