Calculation of percent porosity in a grayscale image

[Seetha S]

Hi,

I'd like to calculate the percent porosity in grayscale images (i.e. get a
numerical estimate of the area covered by pores vs. the area occupied by
polymeric fibers) obtained from x-ray computed tomography to possibly
quantify the overall porosity of the membranes that I've imaged. Could you
please let me know if such a computation is possible in ImageJ?

Thank you.

--
Seetha

[Michael Doube]

Hi Seetha,

You can calculate the volume fraction (or area fraction) using BoneJ, but you have to binarise your image first (e.g. with Image > Adjust > Set Threshold).

http://bonej.org/volumefraction

Michael
________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Seetha S [[hidden email]]
Sent: 24 April 2011 19:45
To: [hidden email]
Subject: Calculation of percent porosity in a grayscale image

Hi,

I'd like to calculate the percent porosity in grayscale images (i.e. get a
numerical estimate of the area covered by pores vs. the area occupied by
polymeric fibers) obtained from x-ray computed tomography to possibly
quantify the overall porosity of the membranes that I've imaged. Could you
please let me know if such a computation is possible in ImageJ?

Thank you.

--
Seetha

[mohebbat mohebbi]

What about measurement of void fraction in 2-d space?

--- On Mon, 4/25/11, Doube, Michael <[hidden email]> wrote:

From: Doube, Michael <[hidden email]>
Subject: Re: Calculation of percent porosity in a grayscale image
To: [hidden email]
Date: Monday, April 25, 2011, 2:16 PM

Hi Seetha,

You can calculate the volume fraction (or area fraction) using BoneJ, but you have to binarise your image first (e.g. with Image > Adjust > Set Threshold).

http://bonej.org/volumefraction

Michael
________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Seetha S [[hidden email]]
Sent: 24 April 2011 19:45
To: [hidden email]
Subject: Calculation of percent porosity in a grayscale image

Hi,

I'd like to calculate the percent porosity in grayscale images (i.e. get a
numerical estimate of the area covered by pores vs. the area occupied by
polymeric fibers) obtained from x-ray computed tomography to possibly
quantify the overall porosity of the membranes that I've imaged. Could you
please let me know if such a computation is possible in ImageJ?

Thank you.

--
Seetha

[Michael Doube]

> What about measurement of void fraction in 2-d space?

Area Fraction is a standard measurement that you can set under Analyze > Set Measurements, and is reported when you hit [M] or run Analyze > Measure

[Seetha S]

Hi Michael,

Thank you for your reply. The images I was referring to are actually 2-D
slices (z-slices) obtained from x-ray computed tomography. I would, however,
also like to measure the void fraction of the 3-D volume that the technique
generates.. but I don't know if it will be possible for me to first load
such an image onto my computer. I haven't done it before and I don't know if
I need a special software to obtain the image from the parent computer.

With regard to my initial query, I tried using area fraction in ImageJ to
obtain the void fraction but I seem to get incorrect estimates from the
software.. even with an almost non-porous membrane slice I got a % Area
value of ~92%. Could you please let me know if I'm doing something wrong
here? I simple opened the image in ImageJ and just hit 'M' after checking
the 'Area fraction' option under Analyze-Set Measurements.

Thank you once again for your time and help.

On Mon, Apr 25, 2011 at 5:46 AM, Doube, Michael <[hidden email]>wrote:

> Hi Seetha,
>
> You can calculate the volume fraction (or area fraction) using BoneJ, but
> you have to binarise your image first (e.g. with Image > Adjust > Set
> Threshold).
>
> http://bonej.org/volumefraction
>
> Michael
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Seetha S [
> [hidden email]]
> Sent: 24 April 2011 19:45
> To: [hidden email]
> Subject: Calculation of percent porosity in a grayscale image
>
> Hi,
>
> I'd like to calculate the percent porosity in grayscale images (i.e. get a
> numerical estimate of the area covered by pores vs. the area occupied by
> polymeric fibers) obtained from x-ray computed tomography to possibly
> quantify the overall porosity of the membranes that I've imaged. Could you
> please let me know if such a computation is possible in ImageJ?
>
> Thank you.
>
> --
> Seetha
>

--
Seetha

[Michael Doube]

> I haven't done it before and I don't know if
> I need a special software to obtain the image from the parent computer.

For list members to  help you, you will need to supply more details:
What kind of images do you have (DICOM, TIFF, BMP ...)?
Who manufactured the scanner?
Can we see some images (post them somewhere and send the URL to this list)?

>With regard to my initial query, I tried using area fraction in ImageJ to
>obtain the void fraction but I seem to get incorrect estimates from the
>software.. even with an almost non-porous membrane slice I got a % Area
>value of ~92%.

Could it be that you have measured the area fraction of non-membrane?
In which case the area fraction of membrane would be ~8%.
The parts you are measuring typically go red when you are thresholding.

Michael

[Seetha S]

Hi Michael,

Sorry for not replying sooner.. I was trying to get the information you
asked for from Xradia Inc. which is the company that sold us the x-ray CT.

What kind of images do you have (DICOM, TIFF, BMP ...)?
I can get the images in apparently any file format from the instrument.
Who manufactured the scanner?
The manufacturer is Xradia Inc.
Can we see some images (post them somewhere and send the URL to this list)?
These first two are SEM images whose %pore area I was trying to get by using
the 'Area fraction' option in ImageJ and the third link is to a x-ray CT
sample report for one of my membranes in a Word file.

https://dropbox.uconn.edu/dropbox?n=COOLTEXX_40001.tif&p=WxLWthqQzzDTnWARc

https://dropbox.uconn.edu/dropbox?n=16-1_20001.tif&p=WbmRdternUuG9bqDV

https://dropbox.uconn.edu/dropbox?n=Neat%20SW30-XLE_Report-2011-04-05.doc&p=W1Lldx2Wvpsul0N12

You mentioned that the area that I am measuring will go red while
thresholding but I don't threshold my images... I wasn't aware that I should
do that before the measurement step. One last question..would I be able to
get the %pore volume (average bulk porosity) of a 3-D volume using BoneJ if
I can figure out a way to load the image onto my laptop.

Thank you very much for your time and help.. I really appreciate it.

On Mon, Apr 25, 2011 at 4:18 PM, Doube, Michael <[hidden email]>wrote:

> > I haven't done it before and I don't know if
> > I need a special software to obtain the image from the parent computer.
>
> For list members to  help you, you will need to supply more details:
> What kind of images do you have (DICOM, TIFF, BMP ...)?
> Who manufactured the scanner?
> Can we see some images (post them somewhere and send the URL to this list)?
>
> >With regard to my initial query, I tried using area fraction in ImageJ to
> >obtain the void fraction but I seem to get incorrect estimates from the
> >software.. even with an almost non-porous membrane slice I got a % Area
> >value of ~92%.
>
> Could it be that you have measured the area fraction of non-membrane?
> In which case the area fraction of membrane would be ~8%.
> The parts you are measuring typically go red when you are thresholding.
>
> Michael
>

--
Seetha

[Michael Schmid]

Hi Seetha,

as far as i can see your sample images are SEM (scanning electron  
microscopy) images, not tomographic images. These images don't  
provide any information about depth (z), so you cannot calculate the  
porosity from them.

Michael
________________________________________________________________

On 29 Apr 2011, at 05:23, Seetha S wrote:

> Hi Michael,
>
> Sorry for not replying sooner.. I was trying to get the information  
> you
> asked for from Xradia Inc. which is the company that sold us the x-
> ray CT.
>
> What kind of images do you have (DICOM, TIFF, BMP ...)?
> I can get the images in apparently any file format from the  
> instrument.
> Who manufactured the scanner?
> The manufacturer is Xradia Inc.
> Can we see some images (post them somewhere and send the URL to  
> this list)?
> These first two are SEM images whose %pore area I was trying to get  
> by using
> the 'Area fraction' option in ImageJ and the third link is to a x-
> ray CT
> sample report for one of my membranes in a Word file.
>
> https://dropbox.uconn.edu/dropbox?
> n=COOLTEXX_40001.tif&p=WxLWthqQzzDTnWARc
>
> https://dropbox.uconn.edu/dropbox?n=16-1_20001.tif&p=WbmRdternUuG9bqDV
>
> https://dropbox.uconn.edu/dropbox?n=Neat%20SW30- 
> XLE_Report-2011-04-05.doc&p=W1Lldx2Wvpsul0N12
>
> You mentioned that the area that I am measuring will go red while
> thresholding but I don't threshold my images... I wasn't aware that  
> I should
> do that before the measurement step. One last question..would I be  
> able to
> get the %pore volume (average bulk porosity) of a 3-D volume using  
> BoneJ if
> I can figure out a way to load the image onto my laptop.
>
> Thank you very much for your time and help.. I really appreciate it.
>
> On Mon, Apr 25, 2011 at 4:18 PM, Doube, Michael  
> <[hidden email]>wrote:
>
>>> I haven't done it before and I don't know if
>>> I need a special software to obtain the image from the parent  
>>> computer.
>>
>> For list members to  help you, you will need to supply more details:
>> What kind of images do you have (DICOM, TIFF, BMP ...)?
>> Who manufactured the scanner?
>> Can we see some images (post them somewhere and send the URL to  
>> this list)?
>>
>>> With regard to my initial query, I tried using area fraction in  
>>> ImageJ to
>>> obtain the void fraction but I seem to get incorrect estimates  
>>> from the
>>> software.. even with an almost non-porous membrane slice I got a  
>>> % Area
>>> value of ~92%.
>>
>> Could it be that you have measured the area fraction of non-membrane?
>> In which case the area fraction of membrane would be ~8%.
>> The parts you are measuring typically go red when you are  
>> thresholding.
>>
>> Michael
>>
>
>
>
> --
> Seetha

[Seetha S]

Would it be possible for me to get the area fraction of these images from
ImageJ? Also, if I can get the 3-D image from the computer hooked up to the
x-ray CT, would it be possible for me to get the bulk porosity using BoneJ?

Thank you!

On Fri, Apr 29, 2011 at 8:31 AM, Michael Schmid <[hidden email]>wrote:

> Hi Seetha,
>
> as far as i can see your sample images are SEM (scanning electron
> microscopy) images, not tomographic images. These images don't provide any
> information about depth (z), so you cannot calculate the porosity from them.
>
> Michael
> ________________________________________________________________
>
>
> On 29 Apr 2011, at 05:23, Seetha S wrote:
>
>  Hi Michael,
>>
>> Sorry for not replying sooner.. I was trying to get the information you
>> asked for from Xradia Inc. which is the company that sold us the x-ray CT.
>>
>> What kind of images do you have (DICOM, TIFF, BMP ...)?
>> I can get the images in apparently any file format from the instrument.
>> Who manufactured the scanner?
>> The manufacturer is Xradia Inc.
>> Can we see some images (post them somewhere and send the URL to this
>> list)?
>> These first two are SEM images whose %pore area I was trying to get by
>> using
>> the 'Area fraction' option in ImageJ and the third link is to a x-ray CT
>> sample report for one of my membranes in a Word file.
>>
>> https://dropbox.uconn.edu/dropbox?n=COOLTEXX_40001.tif&p=WxLWthqQzzDTnWARc
>>
>> https://dropbox.uconn.edu/dropbox?n=16-1_20001.tif&p=WbmRdternUuG9bqDV
>>
>>
>> https://dropbox.uconn.edu/dropbox?n=Neat%20SW30-XLE_Report-2011-04-05.doc&p=W1Lldx2Wvpsul0N12
>>
>> You mentioned that the area that I am measuring will go red while
>> thresholding but I don't threshold my images... I wasn't aware that I
>> should
>> do that before the measurement step. One last question..would I be able to
>> get the %pore volume (average bulk porosity) of a 3-D volume using BoneJ
>> if
>> I can figure out a way to load the image onto my laptop.
>>
>> Thank you very much for your time and help.. I really appreciate it.
>>
>> On Mon, Apr 25, 2011 at 4:18 PM, Doube, Michael <[hidden email]
>> >wrote:
>>
>>  I haven't done it before and I don't know if
>>>> I need a special software to obtain the image from the parent computer.
>>>>
>>>
>>> For list members to  help you, you will need to supply more details:
>>> What kind of images do you have (DICOM, TIFF, BMP ...)?
>>> Who manufactured the scanner?
>>> Can we see some images (post them somewhere and send the URL to this
>>> list)?
>>>
>>>  With regard to my initial query, I tried using area fraction in ImageJ
>>>> to
>>>> obtain the void fraction but I seem to get incorrect estimates from the
>>>> software.. even with an almost non-porous membrane slice I got a % Area
>>>> value of ~92%.
>>>>
>>>
>>> Could it be that you have measured the area fraction of non-membrane?
>>> In which case the area fraction of membrane would be ~8%.
>>> The parts you are measuring typically go red when you are thresholding.
>>>
>>> Michael
>>>
>>>
>>
>>
>> --
>> Seetha
>>
>

--
Seetha

[mj_metallurgist]

I am doing my PROJECT of masters and have to submit on friday please urgently i need to know two things how can i calculate grain size and porosity in my aluminium sample which i have attached as well ???


I am also pasting here the method by which i am trying to calculate the grain size please if i can know it is correct or not ??

open image
2. select 8 bit
3. set scale
4. crop image and then save as jpeg (the extension should be .tiff)
5. threshold image (B&W)
6. erode
7. dilate
8. analyze particles

RESULTS shown in some units ... are these units the same i set in scale i mean if i set scale in MICRON the results are in micron or in pixels ???

pelase any one  can tell me porosity measurements as well .. much thanks !

[t-ship]

I recently did a study of aluminium welds, looking for porosity distribution and total volume area.
I applied a threshold
Analyze-Analyzed particles
  size 0-infinity, Circularity .85-1.0   show masks (needs binarized image)
Then..3 object counter
Subtract pore total volume from total volume

my stacks were 512X512X250 slices