Thread Subject: Related to image intensity

Hi,
I am developing a code with Labview for a camera in my lab.
matlab is also involved for post processing of data
collected. I have a doubt regarding the matlab part. I
collect images of a 9 well slide which has flourescent dyes,
these emit flouresence at a particular wavelength. The wells
itself are small compared to the total area of the slide. I
want to learn if there willbe any effect in the intensities
due to light reflected from the slide. Please help me with
this. Find below the link for image to help you better
understand my problem.

http://www.picoodle.com/view.php?img=/4/5/30/f_SlideswithYm_1339a7b.jpg&srv=img29

In article <g1pkht$70l$1@fred.mathworks.com>,
subash <subashch@gmail.com> wrote:

>collect images of a 9 well slide which has flourescent dyes,
>these emit flouresence at a particular wavelength. The wells
>itself are small compared to the total area of the slide. I
>want to learn if there willbe any effect in the intensities
>due to light reflected from the slide.

That would, I think, depend upon how you took the picture. If you
set the apeture and focus, and then let the camera choose the
exposure time, then Yes, the intensity (and possible saturation)
of the images within the wells will be affected by the timing the
camera meters off as appropriate. Different cameras meter the
light in different ways -- single center spot reading, or
average four readings from a square near the center are common,
and I believe I encountered one that averaged from six spots near
the center arranged in a hexagon.

If you set the apeture and focus and timing yourself, you would
presumably take care to meter off of the wells (or based upon
historical maximums of the brightness the wells can reach).


Or am I misunderstanding the question??
--
  "I want to be remembered as the guy who gave his all whenever
  he was on the field." -- Walter Payton

On May 30, 3:29=A0pm, "subash " <subas...@gmail.com> wrote:
> Hi,
> I am developing a code with Labview for a camera in my lab.
> matlab is also involved for post processing of data
> collected. I have a doubt regarding the matlab part. I
> collect images of a 9 well slide which has flourescent dyes,
> these emit flouresence at a particular wavelength. The wells
> itself are small compared to the total area of the slide. I
> want to learn if there willbe any effect in the intensities
> due to light reflected from the slide. Please help me with
> this. Find below the link for image to help you better
> understand my problem.
>
> http://www.picoodle.com/view.php?img=3D/4/5/30/f_SlideswithYm_1339a7b.j...=


--------------------------------------------------------
subash:
Good question, and thanks for posting an image. But what a crummy
image! There are several problems with this that need correcting, or
maybe I just don't understand your situation. For starters, the easy
one: your focal length is wrong. You have too short a focal length on
your lens. Why do I say this? It looks like you are looking at the
wells not at normal incidence because I can tell by looking at the
shadows of the slide sides on the substance in the wells. You need to
use a longer focal length lens and back up some (get farther away from
your slide) so that you're not looking at an angle into any of the
wells but are looking straight down them.

Next, I see what appears to be several specular reflections
(reflections of your light source). If this were fluorescence imaging
you would not see the reflections of your light source because your
excitation filter would block them.

OK, well let's assume you're using just regular illumination without
an excitation filter, you could block most of the specular reflections
using a polarizer over your light source and a cross polarizer over
your lens. This is a pretty standard method for knocking out specular
reflections and works well for even non-fluorescent imaging. However,
if it's truly fluorescent, you'd be better off using an excitation
filter in front of your lens than using polarizers because it will let
more light through. Also, specular reflections like these can cause
lens flare (additional copies of the reflection caused by internal
reflection off the multiple internal lens surfaces - like you see when
you point a camera at the sun), so it's good to get rid of them lest
they cause spurious signals at undesired locations.

OK, now suppose you fixed those problems and have a better image. I
do not see much evidence that there is light reflected off the sides
of the well. Moreover, who cares? Don't you want to measure ALL
light emitted by the substance. Indeed, even if ALL the light were
reflected, you probably would even be better off. In fact if your
wells were perfectly reflecting little silvered mirror-like cups, that
would be best wouldn't it? You WANT to collect ALL the light being
emitted so if it bounces off the sides, fine - that's good. On the
other hand, what if the walls were perfectly absorbing black so that
no light at all reflected off? Well, your total light collected would
be less and your signal to noise ratio would be less but nonetheless,
you could still compare the wells to each other based on the intensity
collected from each well. So, bottom line, I don't think the amount
of light reflecting off the well sides will hurt you in comparing the
well intensities. It could be a lot or a little but you can still
compare.

Finally, the worst case scenario: you can't correct the imaging setup
for some reason, and you're stuck with the bad images. The specular
reflections (bright spots) all seem brighter than the substance. I'd
threshold the bright areas and replace them with the surround. MATLAB
has (in the image processing toolbox), if I recall correctly, a
function that can do a hole fill on a gray scale image by filling the
hole pixels with values that are the distance-weighted sum of the edge
of the hole-edge values. I'd fill in the holes that way, then just
integrate (sum up) all the gray levels in each well. Then you can
compare them. You can't just exclude the bright areas because the
area of the bright spots is different from well to well. But like I
said, this is a last resort. However, summing the pixel values in
each well is what you want to do with a good image - you just woudn't
have to do the bright spot thresholding stuff.
Regards,
ImageAnalyst

On May 30, 11:06=A0pm, ImageAnalyst <imageanal...@mailinator.com> wrote:
> On May 30, 3:29=A0pm, "subash " <subas...@gmail.com> wrote:
>
> > Hi,
> > I am developing a code with Labview for a camera in my lab.
> > matlab is also involved for post processing of data
> > collected. I have a doubt regarding the matlab part. I
> > collect images of a 9 well slide which has flourescent dyes,
> > these emit flouresence at a particular wavelength. The wells
> > itself are small compared to the total area of the slide. I
> > want to learn if there willbe any effect in the intensities
> > due to light reflected from the slide. Please help me with
> > this. Find below the link for image to help you better
> > understand my problem.
>
> >http://www.picoodle.com/view.php?img=3D/4/5/30/f_SlideswithYm_1339a7b.j..=
.
>
> --------------------------------------------------------
> subash:
> Good question, and thanks for posting an image. =A0But what a crummy
> image! There are several problems with this that need correcting, or
> maybe I just don't understand your situation. =A0For starters, the easy
> one: your focal length is wrong. =A0You have too short a focal length on
> your lens. =A0Why do I say this? =A0It looks like you are looking at the
> wells not at normal incidence because I can tell by looking at the
> shadows of the slide sides on the substance in the wells. =A0You need to
> use a longer focal length lens and back up some (get farther away from
> your slide) so that you're not looking at an angle into any of the
> wells but are looking straight down them.
>
> Next, I see what appears to be several specular reflections
> (reflections of your light source). =A0If this were fluorescence imaging
> you would not see the reflections of your light source because your
> excitation filter would block them.
>
> OK, well let's assume you're using just regular illumination without
> an excitation filter, you could block most of the specular reflections
> using a polarizer over your light source and a cross polarizer over
> your lens. =A0This is a pretty standard method for knocking out specular
> reflections and works well for even non-fluorescent imaging. =A0However,
> if it's truly fluorescent, you'd be better off using an excitation
> filter in front of your lens than using polarizers because it will let
> more light through. =A0Also, specular reflections like these can cause
> lens flare (additional copies of the reflection caused by internal
> reflection off the multiple internal lens surfaces - like you see when
> you point a camera at the sun), so it's good to get rid of them lest
> they cause spurious signals at undesired locations.
>
> OK, now suppose you fixed those problems and have a better image. =A0I
> do not see much evidence that there is light reflected off the sides
> of the well. =A0Moreover, who cares? =A0Don't you want to measure ALL
> light emitted by the substance. =A0Indeed, even if ALL the light were
> reflected, you probably would even be better off. =A0In fact if your
> wells were perfectly reflecting little silvered mirror-like cups, that
> would be best wouldn't it? =A0You WANT to collect ALL the light being
> emitted so if it bounces off the sides, fine - that's good. =A0On the
> other hand, what if the walls were perfectly absorbing black so that
> no light at all reflected off? =A0Well, your total light collected would
> be less and your signal to noise ratio would be less but nonetheless,
> you could still compare the wells to each other based on the intensity
> collected from each well. =A0So, bottom line, I don't think the amount
> of light reflecting off the well sides will hurt you in comparing the
> well intensities. =A0It could be a lot or a little but you can still
> compare.
>
> Finally, the worst case scenario: you can't correct the imaging setup
> for some reason, and you're stuck with the bad images. =A0The specular
> reflections (bright spots) all seem brighter than the substance. =A0I'd
> threshold the bright areas and replace them with the surround. =A0MATLAB
> has (in the image processing toolbox), if I recall correctly, a
> function that can do a hole fill on a gray scale image by filling the
> hole pixels with values that are the distance-weighted sum of the edge
> of the hole-edge values. =A0I'd fill in the holes that way, then just
> integrate (sum up) all the gray levels in each well. =A0Then you can
> compare them. =A0You can't just exclude the bright areas because the
> area of the bright spots is different from well to well. =A0But like I
> said, this is a last resort. =A0However, summing the pixel values in
> each well is what you want to do with a good image - you just woudn't
> have to do the bright spot thresholding stuff.
> Regards,
> ImageAnalyst

----------------------------------------------------------
And just to second what Walter said (and I forgot to say above), you
need to have your camera in manual mode, not auto mode, so that
intensities can be compared from image to image. You don't want to
have the same signal only to find out that the camera reduced the
exposure because the scene was brighter, and now you're not measuring
that increased brightness like you want. Also, for best correction,
you'd want to have some kind of standard in there. Spectralon is a
good standard to use since it's reflectance is fairly flat and high
well down into the UV. You can measure the Spectralon signal and
correct each image back to one image that you define as your "master"
image.
Regards,
ImageAnalyst

ImageAnalyst <imageanalyst@mailinator.com> wrote in message
<3ec170aa-49a2-4112-9a3d-d6cb3767c381@59g2000hsb.googlegroups.com>...
> On May 30, 11:06=A0pm, ImageAnalyst
<imageanal...@mailinator.com> wrote:
> > On May 30, 3:29=A0pm, "subash " <subas...@gmail.com> wrote:
> >
> > > Hi,
> > > I am developing a code with Labview for a camera in my
lab.
> > > matlab is also involved for post processing of data
> > > collected. I have a doubt regarding the matlab part. I
> > > collect images of a 9 well slide which has flourescent
dyes,
> > > these emit flouresence at a particular wavelength. The
wells
> > > itself are small compared to the total area of the
slide. I
> > > want to learn if there willbe any effect in the
intensities
> > > due to light reflected from the slide. Please help me with
> > > this. Find below the link for image to help you better
> > > understand my problem.
> >
> >
>http://www.picoodle.com/view.php?img=3D/4/5/30/f_SlideswithYm_1339a7b.j..=
> .
> >
> > --------------------------------------------------------
> > subash:
> > Good question, and thanks for posting an image. =A0But
what a crummy
> > image! There are several problems with this that need
correcting, or
> > maybe I just don't understand your situation. =A0For
starters, the easy
> > one: your focal length is wrong. =A0You have too short a
focal length on
> > your lens. =A0Why do I say this? =A0It looks like you
are looking at the
> > wells not at normal incidence because I can tell by
looking at the
> > shadows of the slide sides on the substance in the
wells. =A0You need to
> > use a longer focal length lens and back up some (get
farther away from
> > your slide) so that you're not looking at an angle into
any of the
> > wells but are looking straight down them.
> >
> > Next, I see what appears to be several specular reflections
> > (reflections of your light source). =A0If this were
fluorescence imaging
> > you would not see the reflections of your light source
because your
> > excitation filter would block them.
> >
> > OK, well let's assume you're using just regular
illumination without
> > an excitation filter, you could block most of the
specular reflections
> > using a polarizer over your light source and a cross
polarizer over
> > your lens. =A0This is a pretty standard method for
knocking out specular
> > reflections and works well for even non-fluorescent
imaging. =A0However,
> > if it's truly fluorescent, you'd be better off using an
excitation
> > filter in front of your lens than using polarizers
because it will let
> > more light through. =A0Also, specular reflections like
these can cause
> > lens flare (additional copies of the reflection caused
by internal
> > reflection off the multiple internal lens surfaces -
like you see when
> > you point a camera at the sun), so it's good to get rid
of them lest
> > they cause spurious signals at undesired locations.
> >
> > OK, now suppose you fixed those problems and have a
better image. =A0I
> > do not see much evidence that there is light reflected
off the sides
> > of the well. =A0Moreover, who cares? =A0Don't you want
to measure ALL
> > light emitted by the substance. =A0Indeed, even if ALL
the light were
> > reflected, you probably would even be better off. =A0In
fact if your
> > wells were perfectly reflecting little silvered
mirror-like cups, that
> > would be best wouldn't it? =A0You WANT to collect ALL
the light being
> > emitted so if it bounces off the sides, fine - that's
good. =A0On the
> > other hand, what if the walls were perfectly absorbing
black so that
> > no light at all reflected off? =A0Well, your total light
collected would
> > be less and your signal to noise ratio would be less but
nonetheless,
> > you could still compare the wells to each other based on
the intensity
> > collected from each well. =A0So, bottom line, I don't
think the amount
> > of light reflecting off the well sides will hurt you in
comparing the
> > well intensities. =A0It could be a lot or a little but
you can still
> > compare.
> >
> > Finally, the worst case scenario: you can't correct the
imaging setup
> > for some reason, and you're stuck with the bad images.
=A0The specular
> > reflections (bright spots) all seem brighter than the
substance. =A0I'd
> > threshold the bright areas and replace them with the
surround. =A0MATLAB
> > has (in the image processing toolbox), if I recall
correctly, a
> > function that can do a hole fill on a gray scale image
by filling the
> > hole pixels with values that are the distance-weighted
sum of the edge
> > of the hole-edge values. =A0I'd fill in the holes that
way, then just
> > integrate (sum up) all the gray levels in each well.
=A0Then you can
> > compare them. =A0You can't just exclude the bright areas
because the
> > area of the bright spots is different from well to well.
=A0But like I
> > said, this is a last resort. =A0However, summing the
pixel values in
> > each well is what you want to do with a good image - you
just woudn't
> > have to do the bright spot thresholding stuff.
> > Regards,
> > ImageAnalyst
>
> ----------------------------------------------------------
> And just to second what Walter said (and I forgot to say
above), you
> need to have your camera in manual mode, not auto mode, so
that
> intensities can be compared from image to image. You
don't want to
> have the same signal only to find out that the camera
reduced the
> exposure because the scene was brighter, and now you're
not measuring
> that increased brightness like you want. Also, for best
correction,
> you'd want to have some kind of standard in there.
Spectralon is a
> good standard to use since it's reflectance is fairly flat
and high
> well down into the UV. You can measure the Spectralon
signal and
> correct each image back to one image that you define as
your "master"
> image.
> Regards,
> ImageAnalyst


Thank you Image Analyst. I too have a feeling about the
focal length, these images were from the first set of my
experiments. I need to change the focal length as you
suggested and about the specular reflections I do use
excitation filter, I don't know if my filters are not
working properly or not. The incident light is not right
over the slides, it is at an angle and thus the shadows. I
tried to position the light source to minimize the shadows,
but since the lens is close to the slides I could not get
them out completely. I will try and correct these problems
and hopefully my images turn out well.

Thank you for your time and helpful suggestion. Do
appreciate it.

Regards,
Subash

roberson@ibd.nrc-cnrc.gc.ca (Walter Roberson) wrote in
message <g1plbu$7dk$1@canopus.cc.umanitoba.ca>...
> In article <g1pkht$70l$1@fred.mathworks.com>,
> subash <subashch@gmail.com> wrote:
>
> >collect images of a 9 well slide which has flourescent dyes,
> >these emit flouresence at a particular wavelength. The wells
> >itself are small compared to the total area of the slide. I
> >want to learn if there willbe any effect in the intensities
> >due to light reflected from the slide.
>
> That would, I think, depend upon how you took the picture.
If you
> set the apeture and focus, and then let the camera choose the
> exposure time, then Yes, the intensity (and possible
saturation)
> of the images within the wells will be affected by the
timing the
> camera meters off as appropriate. Different cameras meter the
> light in different ways -- single center spot reading, or
> average four readings from a square near the center are
common,
> and I believe I encountered one that averaged from six
spots near
> the center arranged in a hexagon.
>
> If you set the apeture and focus and timing yourself, you
would
> presumably take care to meter off of the wells (or based upon
> historical maximums of the brightness the wells can reach).
>
>
> Or am I misunderstanding the question??
> --
> "I want to be remembered as the guy who gave his all
whenever
> he was on the field." -- Walter Payton



Thank you Payton for your reply and that was a good point to
consider. I set all the values manually and I have no idea
how my camera meters. I got a software from a company and
that does not allow me to open any subVIs. I should probably
call them up and find it out. However, I did not understand
the metering part (average four readings from a square,
hexagonal ), could you provide me with some links that can
help me better understand these concepts. I would appreciate
your effort for this.

Regards
Subash

In article <g20tvt$88o$1@fred.mathworks.com>,
subash <subashch@gmail.com> wrote:

>I set all the values manually and I have no idea
>how my camera meters. I got a software from a company and
>that does not allow me to open any subVIs. I should probably
>call them up and find it out. However, I did not understand
>the metering part (average four readings from a square,
>hexagonal ), could you provide me with some links that can
>help me better understand these concepts.

Cameras that choose the shutter speed automatically measure the
light intensity, and choose a shutter speed so that the
light will not oversaturate the CCD (or whatever imaging technology
they are using.) They do *not*, however, read every value in the CCD
matrix and calculate means and standard deviations over the whole
image and ensure that no more than some fraction of the image would
be oversaturated. Instead, they take a shortcut, and only read
out the current values at a limited number of sensor locations.

Cheaper cameras only read off -one- location (generally the one
in the centre of the field) and assume the brightness of everything
else will be more or less the same. In general, this doesn't work
very well, as people seldom centre the field of view directly on
the brightest spot, so on cameras that do only single-spot metering,
it is not uncommon to get images that are overexposed in places.

Most consumer grade digital cameras read off -four- locations
at the corners of a rectangle, generally about 1/4 or 1/3 of the way
out in the field of view (relative to the centre). They then often
average the readings; some of them might also do some processing to
ensure that the shutter time so determined will not result in
oversaturation at any of the four points. [Interestingly, better
consumer-grade cameras often include an option to switch to single-spot
light metering: if that is combined with a settings-lock (e.g., by
holding down the shutter button half-way) then an experienced photographer
can take into account that the lighting in the scene is uneven,
have the camera meter at a point that the photographer knows
(through experience) would be a good light for the scene, and then move
the camera back to the main image location and use the "saved" shutter
speed.]

Some prosumer or professional cameras meter off of more than four
points. I am not positive which brand I encountered this with, but
I think it was Olympus that metered off six points arranged in a hexagon
at about the same distance out from the center that most cameras would
use for the four-spot imaging.


If you are controlling *all* the settings of your camera, then the
number of points that your camera meters off of for its automatic
exposure control (or their location) is not relevant to you, as you
would not be using AES (Automatic Exposure System). But that still
leaves you with the difficulty of metering the light on typical
images so that you get good exposures with whatever shutter-speed
you choose.
--
  "Allegories are in the realm of thoughts, what ruins are in
  the realm of things." -- Walter Benjamin

On Jun 2, 9:47=A0am, "subash " <subas...@gmail.com> wrote:
> ImageAnalyst <imageanal...@mailinator.com> wrote in message
>
> <3ec170aa-49a2-4112-9a3d-d6cb3767c...@59g2000hsb.googlegroups.com>...
>
>
>
> > On May 30, 11:06=3DA0pm, ImageAnalyst
> <imageanal...@mailinator.com> wrote:
> > > On May 30, 3:29=3DA0pm, "subash " <subas...@gmail.com> wrote:
>
> > > > Hi,
> > > > I am developing a code with Labview for a camera in my
> lab.
> > > > matlab is also involved for post processing of data
> > > > collected. I have a doubt regarding the matlab part. I
> > > > collect images of a 9 well slide which has flourescent
> dyes,
> > > > these emit flouresence at a particular wavelength. The
> wells
> > > > itself are small compared to the total area of the
> slide. I
> > > > want to learn if there willbe any effect in the
> intensities
> > > > due to light reflected from the slide. Please help me with
> > > > this. Find below the link for image to help you better
> > > > understand my problem.
>
> >http://www.picoodle.com/view.php?img=3D3D/4/5/30/f_SlideswithYm_1339a7b..=
.
> > .
>
> > > --------------------------------------------------------
> > > subash:
> > > Good question, and thanks for posting an image. =3DA0But
> what a crummy
> > > image! There are several problems with this that need
> correcting, or
> > > maybe I just don't understand your situation. =3DA0For
> starters, the easy
> > > one: your focal length is wrong. =3DA0You have too short a
> focal length on
> > > your lens. =3DA0Why do I say this? =3DA0It looks like you
> are looking at the
> > > wells not at normal incidence because I can tell by
> looking at the
> > > shadows of the slide sides on the substance in the
>
> wells. =3DA0You need to> > use a longer focal length lens and back up some=
 (get
> farther away from
> > > your slide) so that you're not looking at an angle into
> any of the
> > > wells but are looking straight down them.
>
> > > Next, I see what appears to be several specular reflections
> > > (reflections of your light source). =3DA0If this were
>
> fluorescence imaging> > you would not see the reflections of your light so=
urce
> because your
> > > excitation filter would block them.
>
> > > OK, well let's assume you're using just regular
>
> illumination without> > an excitation filter, you could block most of the
>
> specular reflections> > using a polarizer over your light source and a cro=
ss
> polarizer over
> > > your lens. =3DA0This is a pretty standard method for
>
> knocking out specular> > reflections and works well for even non-fluoresce=
nt
>
> imaging. =3DA0However,
>
>
>
> > > if it's truly fluorescent, you'd be better off using an
> excitation
> > > filter in front of your lens than using polarizers
> because it will let
> > > more light through. =3DA0Also, specular reflections like
> these can cause
> > > lens flare (additional copies of the reflection caused
> by internal
> > > reflection off the multiple internal lens surfaces -
> like you see when
> > > you point a camera at the sun), so it's good to get rid
> of them lest
> > > they cause spurious signals at undesired locations.
>
> > > OK, now suppose you fixed those problems and have a
> better image. =3DA0I
> > > do not see much evidence that there is light reflected
> off the sides
> > > of the well. =3DA0Moreover, who cares? =3DA0Don't you want
> to measure ALL
> > > light emitted by the substance. =3DA0Indeed, even if ALL
> the light were
> > > reflected, you probably would even be better off. =3DA0In
> fact if your
> > > wells were perfectly reflecting little silvered
>
> mirror-like cups, that
>
>
>
>
>
> > > would be best wouldn't it? =3DA0You WANT to collect ALL
> the light being
> > > emitted so if it bounces off the sides, fine - that's
> good. =3DA0On the
> > > other hand, what if the walls were perfectly absorbing
> black so that
> > > no light at all reflected off? =3DA0Well, your total light
> collected would
> > > be less and your signal to noise ratio would be less but
> nonetheless,
> > > you could still compare the wells to each other based on
> the intensity
> > > collected from each well. =3DA0So, bottom line, I don't
> think the amount
> > > of light reflecting off the well sides will hurt you in
> comparing the
> > > well intensities. =3DA0It could be a lot or a little but
> you can still
> > > compare.
>
> > > Finally, the worst case scenario: you can't correct the
> imaging setup
> > > for some reason, and you're stuck with the bad images.
> =3DA0The specular
> > > reflections (bright spots) all seem brighter than the
> substance. =3DA0I'd
> > > threshold the bright areas and replace them with the
> surround. =3DA0MATLAB
> > > has (in the image processing toolbox), if I recall
> correctly, a
> > > function that can do a hole fill on a gray scale image
> by filling the
> > > hole pixels with values that are the distance-weighted
> sum of the edge
> > > of the hole-edge values. =3DA0I'd fill in the holes that
> way, then just
> > > integrate (sum up) all the gray levels in each well.
> =3DA0Then you can
> > > compare them. =3DA0You can't just exclude the bright areas
> because the
> > > area of the bright spots is different from well to well.
> =3DA0But like I
> > > said, this is a last resort. =3DA0However, summing the
> pixel values in
> > > each well is what you want to do with a good image - you
> just woudn't
> > > have to do the bright spot thresholding stuff.
> > > Regards,
> > > ImageAnalyst
>
> > ----------------------------------------------------------
> > And just to second what Walter said (and I forgot to say
> above), you
> > need to have your camera in manual mode, not auto mode, so
> that
> > intensities can be compared from image to image. =A0You
> don't want to
> > have the same signal only to find out that the camera
> reduced the
> > exposure because the scene was brighter, and now you're
> not measuring
> > that increased brightness like you want. =A0Also, for best
> correction,
> > you'd want to have some kind of standard in there.
> Spectralon is a
> > good standard to use since it's reflectance is fairly flat
> and high
> > well down into the UV. =A0You can measure the Spectralon
> signal and
> > correct each image back to one image that you define as
> your "master"
> > image.
> > Regards,
> > ImageAnalyst
>
> Thank you Image Analyst. I too have a feeling about the
> focal length, these images were from the first set of my
> experiments. I need to change the focal length as you
> suggested and about the specular reflections I do use
> excitation filter, I don't know if my filters are not
> working properly or not. The incident light is not right
> over the slides, it is at an angle and thus the shadows. I
> tried to position the light source to minimize the shadows,
> but since the lens is close to the slides I could not get
> them out completely. I will try and correct these problems
> and hopefully my images turn out well.
>
> Thank you for your time and helpful suggestion. Do
> appreciate it.
>
> Regards,
> Subash- Hide quoted text -
>
> - Show quoted text -- Hide quoted text -
>
> - Show quoted text -- Hide quoted text -
>
> - Show quoted text -

----------------------------------------------------------------------------=

Subash:
The best sort of light would be broad diffuse light so that shadows
are minimized. That might be tough if you're using a flash lamp for
excitation rather than black light fluorescent tubes, which leads to
the second recommendation. Enclose your plates in a box which is
painted with reflective paint on the interior - for example "powder
coated" paint (Google it). This will give a very diffuse reflectance
from virtually every point inside your light booth. Plus, don't have
your light shine directly on your sample, it should be indirect,
preferably behind a baffle, so that the light gets reflected from all
over - every point on the inside of the light booth. This will
minimize specular reflections (hot spots). This is the concept of an
"integrating sphere" (Google it). Unless you're trying to accentuate
shadows (like you're trying to analyze texture, wrinkles, ridges,
etc.) then an integrating sphere-like light booth is a good kind to
have. So even if you're using a small point source of light, such as
a flash lamp (strobe), you can minimize hot spots by hiding it behind
a baffle in an integrating sphere.
Regards,
ImageAnalyst