PoroMedia Help
TOPICS:
Sample preparation for Porosity SSA, 3D Structure Data,
Tortuosity, and Random Walk.
Porosity
SSA and 3D Structure Data guide.
Sample preparation for Surface Structure Analysis.
Horizontal Slice preparation (for Porosity SSA, 3D
Structure Data, Tortuosity, Random Walk):
Below is a sample
of a Ò.bmpÓ image file used in Porosity SSA, 3D Structure Data, Tortuosity and
Random Walk analysis. This
represents one layer of many required to reconstruct a porous 3D
structure. From a given sample set
(example: oak_000 to oak_100), it
is recommended to use as many slices as possible. Often, the first few and last layers are
not ideal. Find the subset that represents the largest consistent ÒcubeÓ of the
sample (perhaps oak_025 to oak_075).
There are several
required coordinates that need to be extracted from this image as input to the
desired analysis. This can be done
using any image editing software (such as Microsoft Paint, or Adobe
Photoshop). Keep in mind that these
coordinates need to be valid for all slices of this sample (it is recommended
to look at the first slice and last slice, and confirm the coordinates are
valid for both). The black square
outline in the above image represents an area that is valid in all other slices
that are part of this sample.
First, find the
coordinates that represent the top-left most valid location of the sample. In
the above image, these coordinates are represented in yellow (100,100). These are the X and Y start points.
Next, find the
width of the valid area. The above
image represents this in green (300).
Finally, find the
height of the valid area. Seen
above in blue (200).
(Hint: If you use MS Paint, simply drag a
rectangle from the start point to see a (width, height) set).
The following is a
more detailed description for each type of analysis.
Porosity
SSA and 3D Structure Data:
Below is the setup
dialog for Porosity SSA and 3D Structure Data analysis (3D Structure Data becomes
available after Porosity SSA is ran). It is accessed through Analysis
-> Porosity SSAÉ menu.
Labels 1 through 13 are the same for all horizontal slice analysis
(SSA, 3D Structure, Tortuosity, and Random Walk). Here is a description of each:
1)
Click
to pick the directory with sample images.
2)
Displays
directory, can copy/paste into this field.
3)
Enter
the constant part of all image files. For example, if a sample set contains
images oak_025.bmp tooak_075.bmp, Òoak_0Ó must be entered into this field.
4)
Enter
a distinguishable name for this run.
Sample Size:
5)
Enter
the first slice number (the number that is attached to the constant part of
image file). Continuing on the
previous example, Ò25Ó would be entered here (from oak_025.bmp).
6)
Enter
the last slice number. Continuing
on the previous example, Ò75Ó would be entered here (from oak_075.bmp).
7)
Enter the start of X-coordinate (description).
8)
Enter the start of Y-coordinate (description).
9)
Enter Width (description).
10)
Enter Height (description).
Threshold:
11)
Check this box if the image is inverted (pores are white, fiber is
black). This
is an inverted image.
12)
Check this box to enable field 13 for entering a desired threshold. If this box is not checked, the
threshold is automatically computed as part of the run. It may sometimes be necessary to use a
manual threshold for certain samples for desired results.
13)
Enter a threshold if field 12 is checked.
The threshold determines how the Ò.bmpÓ images are converted to binary
form. A brief explanation is as
follows. The Ò.bmpÓ images are
first converted to raw form, with each pixel having intensity between 0 and
255. Then, any value greater than
the threshold value becomes a binary Ô1Õ (fiber), and any value less than the
threshold value becomes a binary Ô0Õ (pore). If there are problems such as Ò.bmpÓ
images being converted to an all black binary files, the threshold should be
raised to a higher value, and vice versa.
This
is the only unique field for SSA/3D Structure.
14)
Enter the pixel size (in micrometers) of the image file.
Press
ÒOKÓ to start the run, the window will be minimized until the run is
complete. After it is complete 3D
Structure Data becomes available in the Analysis menu. Select it to finish the analysis.
Below is the setup
dialog for Tortuosity analysis. It is accessed through Analysis -> Tortuosity
menu.
Description
of fields 1 through 13 can be found here. There are two different directions for
tortuosity analysis. Transverse
sends a tracer from the top layer to the bottom, while Inplane sends a tracer
from the side of the sample, to a specified distance inside.
Transverse Tortuosity:
14)
Enter the number of tracers to run through the sample in
transverse direction. Entering a
higher number results in a linearly longer run time, but higher accuracy. 1000 runs is a good balance.
15)
Check this box to run Transverse Tortuosity.
Inplane Tortuosity:
16)
This specifies the limit of pixel distance for inplane
tracers. Setting this value too
high can result in very long run times.
17)
Similar to 14, but
for Inplane Tortuosity only.
18)
Check this box to run Inplane Tortuosity.
Advanced:
19)
This determines how many ÒbucketsÓ are in the final tracer
distribution results.
20)
Saves the converted raw image files for examination/debugging.
21)
Saves the converted binary image files for examination/debugging.
Press
ÒOKÓ to start the run, the window will be minimized until the run is complete.
Below is the setup
dialog for Random Walk analysis. It is accessed through Analysis -> Random
Walk menu.
Description
of fields 1 through 13 can be found here. There are two different directions for
tortuosity analysis. Transverse
sends a tracer from the top layer to the bottom, while Inplane sends a tracer
from the side of the sample, to a specified distance inside.
Random Walk Settings:
14)
Enter the number of molecules to diffuse.
15)
Set the limit for the distance (in pixels) that a particle can
travel. This prevents an infinite
run if the particle is in a chamber.
16)
How great of a distance to travel before a data point is recorded.
Advanced Settings:
17)
Enter the pixel size in microns.
18)
Option for saving converted raw image files.
19)
Option for saving converted Binary image files.
Press
ÒOKÓ to start the run, the window will be minimized until the run is complete.
Vertical Slice preparation (for Surface Structure
Analysis):
Below is a sample
of a Ò.bmpÓ image file used in Surface Structure analysis. This represents one slice of many
required to reconstruct a porous 3D structure. Opposed to the horizontal slices, the
vertical one is the side view of a sample, and therefore often requires many
more slices to get an adequate sample.
From a given sample set (example:
oak_0000 to oak_1000), it is recommended to use as many slices as
possible. Find the subset that
represents the largest consistent ÒcubeÓ of the sample (perhaps oak_0400 to
oak_0600).
There are several
required coordinates that need to be extracted from this image as input to
Surface Structure Analysis. This
can be done using any image editing software (such as Microsoft Paint, or Adobe
Photoshop). Keep in mind that these
coordinates need to be valid for all slices of this sample (it is recommended
to look at the first slice, middle and last slice, and confirm the coordinates
are valid for all).
First, find the
X-coordinates that represent the left most
valid point and the right most valid
point. These should represent a
section of the sample that is fairly level (samples that are scanned at an
angle other than flat horizontal are not as good).
Next, estimate
the highest point of the sample (Top-Upper Limit). It is beneficial to add a few extra
pixel layers for a safe margin, since other slices might have higher points (it
is impractical to scan all slices looking for highest point). Also, estimate
the lowest point on the top layer of the sample (Top-Lower
Limit).
Finally, repeat
this for the bottom layer of the sample.
The Bottom-Upper Limit and Bottom-Lower Limit represent these estimates for
the above sample.
The following is a
more detailed description of running Surface Structure Analysis.
Below is the
dialog for Surface Structure analysis. It is accessed through Analysis
-> Surface Structure menu.
1)
Click
to pick the directory with sample images.
2)
Displays
image directory, can copy/paste.
3)
Enter
the constant part of all image files. For example, if a sample set contains
imagesvert_oak_0200.bmp to vert_oak_0300.bmp, Òvert_oak_0Ó must be entered into
this field.
4)
Click
to pick the directory for run results.
5)
Displays
result directory, can copy/paste.
Sample Properties:
6)
Enter
the first slice number (the number that is attached to the constant part of
image file). Continuing on the
previous example, Ò200Ó would be entered here (from vert_oak_0200.bmp).
7)
Enter
the last slice number. Continuing
on the previous example, Ò300Ó would be entered here (from vert_oak_0300.bmp).
8)
Left end limit (description).
9)
Right end limit (description).
10)
Upper top end limit (description).
11)
Lower top end limit (description).
12)
Upper bottom end limit (description).
13)
Lower bottom end limit (description).
Threshold:
14)
Check this box if the image is inverted (pores are white, fiber is
black). This
is an inverted image.
15)
Check this box to enable field 16 for entering a desired threshold. If this box is not checked, the
threshold is automatically computed as part of the run. It may sometimes be necessary to use a
manual threshold for certain samples for desired results.
16)
Enter a threshold if field 15 is checked.
The threshold determines how the Ò.bmpÓ images are converted to binary
form. A brief explanation is as
follows. The Ò.bmpÓ images are
first converted to raw form, with each pixel having intensity between 0 and
255. Then, any value greater than
the threshold value becomes a binary Ô1Õ (fiber), and any value less than the
threshold value becomes a binary Ô0Õ (pore). If there are problems such as Ò.bmpÓ
images being converted to an all black binary files, the threshold should be
raised to a higher value, and vice versa.
Run Settings:
17)
Run thickness distribution analysis. Result
description.
18)
Run surface volume distribution analysis. Result
description.
19)
Run contact fraction and surface pit distribution analysis for the
top layer. Result
description.
20)
Run contact fraction and surface pit distribution analysis for the
bottom layer. Result
description.
Advanced:
21)
This determines how many ÒbucketsÓ are in the final surface pit distribution
result.
Press
ÒOKÓ to start the run, the window will be minimized until the run is complete.
Surface Structure Analysis Output Files:
All
output files will be in the user specified Results Directory (5).
Depending on what Runs are specified, the following outputs files are
possible. Any Ò.csvÓ
file can be examined using Microsoft Excel.
Run thickness
distribution analysis is checked:
thickness_distribution.csv: Contains
the sample thickness distribution results, and all of the specific run
parameters.
resultThickness: This
directory contains the raw thickness data for every sample slice.
Run surface
volume distribution analysis is checked:
surface_volume_distribution.csv: Contains the sample volume
distribution results, and all of the specific run parameters.
resultSurfaceVolume: This
directory contains the raw data of surface volume run.
Run contact
fraction for Top
is checked:
contact_fraction_Top.csv:
surfacepit_distribution_Top.csv:
resultContactFractTop:
Run contact
fraction for Bottom is checked:
contact_fraction_Bottom.csv:
surfacepit_distribution_Bottom.csv:
resultContactFractBottom: