Image Analysis: Basic Procedures for Description of Plant Structures - Plant Cell Morphogenesis: Methods and Protocols - page 70

{0, 1, 2, …, b −1}, where a is a distance in x -axis direction and
b is a distance in y -axis direction to determine the position of
the initial point of the superimposed point grid (square or
rectangular), the points of which denote the same corner of a
sampled segment (for details see refs. [ 5 , 7 - 9 ]).
6. Apply the principle of SUR sampling design on all hierarchical
levels of your design, i.e., segments, sections, sampling win-
dows, or measured objects.
7. Calculate variability from the results obtained in the pilot
study. The convenient measure of how precise the estimate is
is called the coeffi cient of error , or CE . The coeffi cient of error
is a statistical value used extensively in the stereological litera-
ture (defi ned as the standard deviation divided by mean of the
sample). Coeffi cient of error depends on the number of sam-
ples and corresponds to proportional variability of the esti-
mate; its value should be lower than 0.05. If CE is higher,
modify the sampling design by adding more segments, sec-
tions, sampling windows, or measured objects.
8. In the majority of cases it is possible to apply just a few basic
rules: (1) Five individuals per group are usually enough for the
parameter estimation [ 12 ]; (2) when sampling sections along
the longitudinal axis, cover the whole object systematically, in
such a manner, that 5-10 segments or sections per organ are
sampled; (3) when using sampling windows, 5-10 of them
should be superimposed on each section; and (4) in most cases
it is not necessary to count more than 200 points or intersec-
tions with stereological probe per organ in each compartment
of interest [ 11 ].

3.2 Calibration:
Manual Setting of
Microscopic Image
Calibration

Correct calibration is a must to measure absolute values of a mea-
sured parameter in real units. The calibration is specifi c for particu-
lar acquisition settings. Objective magnifi cation, tubus factor,
camera model (size and resolution of chip), and settings of pixel
binning on camera are particularly important while acquiring the
image with microscope. Calibrate the pixel size with stage microm-
eter or reticle of known dimensions as follows:

1. Acquire the image of stage micrometer or reticle with particular
optical setup (objective and other components affecting
magnifi cation).
2. Open the image of stage micrometer acquired with the same
objective (and other acquisition settings such as image size in
pixels and zoom) as the image (or images) you would like to
analyze. For the instructions to manual setting of known cali-
bration; see Note 6 .
3. Draw a line (Tools/Straight line) along the side of a line whose
size is known. Consider the thickness of the lines of the stage
micrometer—include it only once (Fig. 2 ).