GETOBJ3D
Section: User Commands (1)
Updated: 1 September 1990
Index
Return to Main Contents
NAME
getobj3d - select objects out of an image (3D version)
SYNOPSIS
getobj3d
[-t NN][-b NN][-m NN][-g N][-c NN NN][-s][-a][-o][-v][-f fname] < inseq > outseq
DESCRIPTION
getobj3d
selects an arbitrary number of distinct objects from an
image, and sets the remaining image to a uniform background value
(usually zero). This can be used when only certain objects are
of interest, or to eliminate background noise
around a central object. The program can also be helpful in that
by making the entire background a constant value, data
compression programs will perform better.
This is a 3D version of
getobj.
This program
works on HIPS formatted images of type int, short, and byte.
OPTIONS
The following options are available:
- -t threshold value
-
All grey level values below this value are discarded. Default value is 50.
- -b background value
-
All locations not determined to be part of an object are set
to this value. Default value is 0.
- -m minimum size object
-
Any object smaller than this size is ignored. Usually used in conjunction
with the -a option. (default = 10) NOTE: voxels which fall on a surface
between object and background are not counted in the object size.
- -g N
-
N is the type of connectivity bridge. 0 = no checking (default), 1 =
check for 2D bridges, 2 = check for weak 3D bridges, 3 = check for strong
3D bridges.
Bridges are defined as follows:
2D bridges: in the 2D slice, objects must be connected by at least 3 voxels
3D weak bridges: in 3D, objects must be connected by at least 3 voxels
3D strong bridges: in 3D, objects must be connected by at least 5 voxels
- -d
-
Display statistics (object area, total intensity,
and average intensity) on each object.
- -a
-
Find all objects in the data set.
- -o
-
Output is binary. All pixels in objects are set to 255, and all other pixels are
set to 0.
- -v
-
Verbose mode: displays information along the way. Useful on slow machines
to know the current status of the program.
- -c row column
-
The row, column point which is inside the desired object. If this
point is not in an object, the program exits. If the point is not
specified,
getobj3d
uses the first object it finds from the center of the image.
- -f value file
-
When this argument is given, (col, row, frame) values of object locations
are read from this file. The format of the file should be an
int telling the number of points in the file, followed by a list
of points. For example:
"2
100 120 12
200 210 13"
means use 2 seed values,
one at (100,120,12), and the other at (200, 210,13).
This option over-rides the -c option.
AUTHOR
Brian Tierney, LBL email: bltierney@lbl.gov
SEE ALSO
getobj(1)
DIAGNOSTICS
getobj3d
will give unpredictable results if the objects are not
cleanly separated. If an object is not found at the seed value
specified, it will look for objects within 10 voxels of the
given seed value. Otherwise an error message is printed.
This program may run out of memory if the object size is too large
or if it is given a non-binary image.
BUGS
Report to Brian Tierney.
Index
- NAME
-
- SYNOPSIS
-
- DESCRIPTION
-
- OPTIONS
-
- AUTHOR
-
- SEE ALSO
-
- DIAGNOSTICS
-
- BUGS
-
This document was created by
man2html,
using the manual pages.
Time: 01:45:25 GMT, December 05, 2024