PLEASE NOTE : This is the documentation for the avs module executable
input_create2D, which contains the following modules:
vcustom
vgauss
vgbox
vgcirc
vgconst
vgfractal
vggauss
vgpwl
vgshot
vgsin
vimpulse
vmarr
Any mention of xvimage is actually a "field 2D". Also, the INPUTs and
OUTPUTs, which are mapped to avs parameters, inputs, and outputs, are for the
khoros library routine.
*********************************************************************************
Documentation for avs module vcustom
INPUT
image pointer to xvimage structure pointer
rows number of rows of image to generate
cols number of columns of image to generate
type data type of image to generate
OUTPUT
image output xvimage structure.
DESCRIPTION
vcustom allows the user to create a custom kernel file in
the VIFF format.
An xvimage structure pointer, image, must be passed in to
vcustom. The parameters of image_type, row_size and
col_size are used to determine the characteristics of the
created kernel.
vcustom returns a zero if failure and a one upon success.
SEE ALSO
vcustom(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
vcustom can only create byte, short, int and float data
storage types.
*** vcustom will not work if avs is running in cli mode !!
AUTHOR
Richard Krukar
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgauss
INPUT
OUTPUT
DESCRIPTION
vgauss generates and image that contains one or more Gaus-
sian distributions. The bivariate mean and variance can be
specified along with a (possibly normalized) amplitude and
rotation angle between the major and minor axes of the iso-
contour ellipse. These parameters can be specified on the
command line when only a single Gaussian is desired in the
output image. When multiple Gaussians are desired, the
parameters are specified in a multiband image where the
number of vectors in the image controls the number of Gaus-
sians to be generated.
For the multiple Gaussian case the parameters image should
contain six bands: 1) X peak location, 2) Y peak location,
3) X variance, 4) Y variance, 5) Rotation Angle, 6) Peak
amplitude. All six bands must be present and be in the above
order.
If the normalize flag (-n) is set then the amplitude infor-
mation (either from the command line of the parameter image)
is ignored, and the proper amplitude is computed so that the
Gaussian distribution is a probability distribution.
The rotation angle argument controls the correlation between
the variables in the distribution. The correleation coeffi-
cient is the sine of this angle (in degrees). This angle
must lie in the range of -90 < 0 < 90. Note that +/- 90
degrees are EXCLUDED from the range since this is a degen-
erate case.
If the parameter image is supplied then the values it con-
tains, overrides all other distribution parameters except
the normalize flag.
SEE ALSO
vgauss(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
The output image is of type FLOAT.
AUTHOR
Scott Wilson
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgbox
INPUT
n_cols unsigned short number of cols (x)
n_rows unsigned short number of rows (y)
px_x_w float active area pixel width in
the pixel_cell
px_y_h float active area pixel height in
the pixel_cell
px_x_of float active pixel offset from left
pixel_cell
px_y_of float active pixel offset from top
pixel_cell
grey_out unsigned char grey level outside of paral-
lelogram
grey_in unsigned char grey level inside of paral-
lelogram
im_x_w float parallelogram euclidian width
(in pixel unit)
im_y_h float parallelogram euclidian
height (in pixel unit)
im_x_of float parallelogram: x right offset
(origine: parallelogram center = CCD center)
im_y_of float parallelogram: y downward
offset (origine: parallelogram center = CCD
center)
phi float clockwise parallelogram rota-
tion (radians, rot. center = parallelogram
center)
psi float first parallelogram corner
angle (radians)
OUTPUT
pt_image_data unsigned char a pointer to the memory allo-
cated by the calling routine. holds the
result of the image data generation
DESCRIPTION
vgbox
This procedure fills a buffer with a grey level represen-
tion of a parallelogram projected on a CCD sensor.
It calculates the grey level of each pixel from the inter-
section of the theorical parallelogram and the rectangular
active area of each sensor-pixel.
The sensor is made of pixels suposed to be all identical,
composed of an active rectangular area in a rectangular
cell.
15 parameters are to be provided:
The buffer,
For the sensor:
its number of rows (=cols_size) and number of columns
(=rows_size)
the size of each pixel active area,
the position of active area on the whole pixel area.
For the parallelogram:
X and Y sizes,
its position (horizontal and vertical shifts),
its rotation (in radians).
its first corner angle (in radians).
The grey levels of full lighted pixels and full dark
pixels.
Return Value: 0 on success,
n=1 on warning,
n=2 on error.
speed: about 0.5 ms /pixel on a Sun 3.
about 0.04 ms /pixel on a Sun 4 sparc station
ipc.
SEE ALSO
vgbox(1), intro(3), vipl(3), verror(3), vutils(3) vgbox(1)
RESTRICTIONS
Pixel width + pixel offset must be egal or less than 1.000.
The CCD cells are supposed square in order to obtain the
asked angles, although the active area can be rectangular.
For a parallelepipedic box, the width and heigth sizes of
the box are the euclidian distances between the parallel
lines.
AUTHOR
Jean-Pierre GUERIN (CEA)
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgcirc
INPUT
image input image structure
rows the desired number of rows in output image
cols the desired number of columns in output image
type the data storage type of the resulting image
bgnd the background intensity
fgnd the forground intensity
x the X coordinate of the center of the circle
y the Y coordinate of the center of the circle
d diameter of circle (in pixels)
OUTPUT
image holds the resulting output image.
DESCRIPTION
vgcirc creates an image of a filled circle. The background
intensity is set by the -b argument. The foreground or cir-
cle intensity is set by the -f argument. The circle can be
placed anywhere, and is NOT required to lie completely
within the image boundaries.
The position of the center of the circle and its diameter
can all be specified as floating point numbers, allowing
sub-pixel positioning of the circle.
SEE ALSO
vgcirc(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
vgcirc will generate only BYTE or FLOAT images.
AUTHOR
Scott Wilson
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgconst
INPUT
Input to vgconst :
image pointer to VIFF structure containing image
data to be processed.
rows number of rows in output image.
cols number of columns in output image.
type Uses the data storage type convension from the
defines in the viff header. VFF_TYP_1_BYTE =
char, VFF_TYP_2_BYTE = short, VFF_TYP_4_BYTE =
int, VFF_TYP_FLOAT = float, VFF_TYP_COMPLEX = com-
plex, VFF_TYP_DOUBLE = double, and
VFF_TYP_DCOMPLEX = dcomplex.
real specifices the constant value (level) for the real
part
imag specifices the constant value for the imaginary
part.
bands defines the number of data bands for the resulting
image.
OUTPUT
Output from vgconst :
image pointer to VIFF structure containing image
data after processing.
Return Value: 1 on success, 0 on failure.
DESCRIPTION
vgconst : creates an image of dimension rows * columns *
bands. The image pixel values are constant. The user can
specify the constant value (level).
This routine will create multiband images and works with
data types BIT, BYTE, SHORT, INT, FLOAT, DOUBLE and COMPLEX,
and DCOMPLEX.
For byte type images:
If the level value specified by user is greater than 255,
the level value will be scaled down to 255. If the level
specified by user is less than 0, the level value will by
scaled up to 0.
The resulting images may be of type byte, short, int or
float, or complex.
SEE ALSO
vgconst(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
does not support transformed images.
AUTHOR
Marcelo Teran, Ramiro Jordan, Tom Sauer
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgfractal
INPUT
image xvimage structure
maxlevel maximum number of recursions
hurst desired hurst coeffient of image
sigma standard deviation of random numbers
seed seed for random number generator
add_flag flag to turn random additions on/off
OUTPUT
image output xvimage structure. Holds the resulting
the fractal image.
Return Value: 1 on success, 0 on failure.
DESCRIPTION
vgfractal creates a fractal image with a desired fractal
dimension. A fractal image is created using the midpoint
displacement algorithm, which uses a recursive generation
technique. Numerous types of fractal images may be created
by varying the input parameters.
The output image, which corresponds to the image argument,
is stored as data type FLOAT, and will be a single band
image. The size of the resulting image will be determined
by the maximum number of recursions, (maxlevel). That is,
the resulting image will be square, with dimensions (2^max-
level +1) by (2^maxlevel +1).
The input parameters include:
(maxlevel) maximum number of recursions,
(hurst) desired hurst coefficient,
(sigma) standard deviation of random numbers,
(seed) random number generator seed,
(coast) coastlines,
(add_flag) random additions.
Descriptions on the use of these parameters follow.
The maximum number of recursions (iterations) is specified
by the "maxlevel" recursive iteration argument. The recur-
sive iteration argument must be an integer ranging from 2 to
10. The size of the image will be a power of two of the
recursive iteration argument. That is, it will create an
image with sizes ranging from (5 by 5) to (1025 by 1025).
The default number of recursions is 6, which will result in
an image of size 65 by 65 pixels.
The hurst coefficient of the generated image may be altered
with the "hurst" argument. This allows the user to change
the fractal dimension of the generated image. Since the
hurst coefficient is a measure of the slope of window size
vs range in pixel values, the valid range for the hurst
coefficient is 0 < H < 1.
The fractal dimension is calculated from the hurst coeffi-
cient as follows:
D = 3 - H
where,
D is the fractal dimension
H is the hurst coefficient.
The parameter 3 reflects the 3-dimensional Euclidian
geometry of the system (ie. X, Y, and Z or pixel value).
The standard deviation of the randomly generated numbers
used to create the fractal image may be altered using the
"sigma" argument. Increasing the standard deviation will
permit a wider range of values to be generated by the random
number generator. This will have the same effect as scaling
the values that are generated.
A "seed" may be entered to allow the same image to be gen-
erated, or the system time may be used to set the seed
(default).
The coastline argument, "coast" can be toggled on (default
is off), which will effectively generate a "coastline" by
setting all generated values that fall below zero to be
equal to zero. This creates the effect of a body of water
running up against a landform, or a "coastline".
The random additions argument, "add_flag", can be toggled
on, which will allow random elements to be added to each
point, whereas with random additions turned off (default),
random elements are added only to the new midpoints gen-
erated in each stage.
Note that in order to display the image as a three dimen-
sional relief plot, the data can be plotted in Xprism3.
Theory - The midpoint displacement algorithm constructs the
image using a recursive generation technique. This method
works with square lattices of points or pixels. Given a
resolution (w) of such a lattice or grid, another square
grid is obtained with a resolution of w/sqrt(2) by adding
the midpoints of all squares. The oreintation of the new
lattice is rotated by 45 degrees, and adding the midpoints
of all squares yields the next lattice with a resolution of
w/2 and the same orientation as the first. This is illus-
trated as follows:
o x o o o o o x o x o
x x x o x o x
x o x ----> o o o ----> o x o x o
x x x o x o x
o x o o o o o x o x o
where, o = old points
x = new points
The input parameters passed into the routine are:
maxlevel - specifies the maximum number of recursions,
or image size. ie. 2^maxlevel.
hurst - desired hurst coefficient,
sigma - standard deviation of random numbers,
seed - random number generator seed,
coast - coastlines flag, 0 = off, 1 = on.
add_flag - random additions, 0 = off, 1 = on.
SEE ALSO
vgfractal(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
The resulting output image will be of type FLOAT.
AUTHOR
Charlie Gage
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vggauss
INPUT
image pointer to VIFF structure containing image
data to be processed.
rows number of rows in image
cols number of columns in image
bands number of data bands in image
mean mean of Gaussian data
variance variance of Gaussian data
type data type of resulting image. Must be either
VFF_TYP_FLOAT or VFF_TYP_COMPLEX
doadd flag to indicate if data is to be added to
the input image. If this flag is set, then
add the noise to the data that currently sits
in image.
OUTPUT
image pointer to VIFF structure containing image
data after processing.
Return Value: 1 on success, 0 on failure.
DESCRIPTION
creates a two dimensional multi-band Gaussian (noise) image
of dimension rows * columns or adds noise to the optional
input image. The Gaussian image is generated using the Box-
Muller transformation.
SEE ALSO
vggauss(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
can only generate VFF_TYP_FLOAT and VFF_TYP_COMPLEX type
images.
AUTHOR
Ramiro Jordan, Jeremy Worley
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgpwl
INPUT
image resulting image.
rows number of rows in image
cols number of columns in image
fs sampling frequency
minimum minimum value of function
maximum maximum value of function
xperiod X period of the function in seconds
yperiod Y period of the function in seconds
xrise X rise time in seconds
yrise Y rise time in seconds
xfall X fall time in seconds
yfall Y fall time in seconds
xwidth X width of pulse
ywidth Y width of pulse
type data type of output image. This must be
either VFF_TYP_FLOAT or VFF_TYP_COMPLEX.
OUTPUT
image holds the result of the data generation.
return Value: 1 on success, 0 on failure.
DESCRIPTION
creates a two dimensional piecewise linear image. The image
pixel values follow a square waveform. The user can specify
the period, phase, minima and maxima, sampling frequency,
rise and fall times, and image dimensions.
NOTE: According to the Sampling Theorem, the sampling fre-
quency should be at least twice that of the signal frequency
to generate a signal properly.
SEE ALSO
vgpwl(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
can only generate VFF_TYP_FLOAT and VFF_TYP_COMPLEX type
images.
AUTHOR
Jeremy Worley
COPYRIGHT
Copyright 1992, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgshot
INPUT
image pointer to VIFF structure containing image
data to be processed.
percent spikes percentage
OUTPUT
image pointer to VIFF structure containing image
data after processing.
Return Value: 1 on success, 0 on failure.
DESCRIPTION
corrupts an image uniformly with spikes. The percentage of
spikes must be specified by the user. Spike locations are
determined using a uniformly distributed random number gen-
erator.
SEE ALSO
vgshot(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
can work only with data storage type byte images.
AUTHOR
Marcelo Teran, Ramiro Jordan
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vgsin
INPUT
image pointer to VIFF structure containing image
data to be processed.
rows rows is an integer that specifies the number
of rows in the resultant image.
cols cols is an integer that specifies the number
of columns in the resultant image.
bands bands is an integer that specifies the number
of bands in the resultant image.
amp amp is an float that specifies the maximum
amplitude of the signal.
fs fs is a float that specifies the sampling
frequency for signal generation.
fx fx is a float that specifies the frequency of
the signal in the X direction in
radians/pixel.
fy fy is a float that specifies the frequency of
the signal in the Y direction. in
radians/pixel
px px is a float that specifies the phase shift
in the X direction in radians.
py py is a float that specifies the phase shift
in the Y direction in radians.
type type is an int that specifies the output data
type. It must be one of VFF_TYP_FLOAT or
VFF_TYP_COMPLEX.
OUTPUT
image pointer to VIFF structure containing image
data after processing.
Return Value: 1 on success, 0 on failure.
DESCRIPTION
vgsin creates a two dimensional image of dimension rows *
columns. The image pixel values follow a sinusoidal
waveform. The user can specify the frequency, phase, ampli-
tude, sampling frequency and image dimensions.
The spatial frequencies are specified in radians/pixel, and
the phase offsets are specified in radians. The expression
generated is then
sin(fx*x + px + fy*y + py)
where fx and fy are the frequencies in the X and Y direc-
tions respectively, and px an py are the phase shifts in the
X and Y directions respectively.
The phase fronts of the sinusoid are perpedicular to the
direction given by the angle atan2(b,a). For a reference on
this see J. Goodman, "Introduction to Fourier Optics".
The sampling frequency modifies the generated expression as
follows:
sin(fx*x/fs + px + fy*y/fs + py)
which allows one to shift the frequency to match that of an
incoming data stream (in 1D).
The resulting images may be of type FLOAT or COMPLEX. A
complex image is generated using the formula cos[stuff] +
j*sin[stuff].
SEE ALSO
vgsin(1), intro(3), vipl(3), verror(3), vutils(3)
lvgconst(3)
RESTRICTIONS
vgsin can only generate VFF_TYP_FLOAT and VFF_TYP_COMPLEX
type images.
AUTHOR
Jeremy Worley, Marcelo Teran, Ramiro Jordan
COPYRIGHT
Copyright 1992, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vimpulse
INPUT
image pointer to VIFF structure containing image
data to be processed.
deltax impulse spacing in x direction
deltay impulse spacing in y direction
nx number of impulses in x direction
ny number of impulses in y direction
xoff the X offset for the first impulse from the
upper left corner of the image
yoff the Y offset for the first impulse from the
upper left corner of the image
rows number of rows in the image to be generated
cols number of columns in the image to be gen-
erated
OUTPUT
image pointer to VIFF structure containing image
data after processing.
Return Value: 1 on success, 0 on failure.
DESCRIPTION
DIAGNOSTICS
The inputs are:
*image - pointer to xvimage structure pointer to receive
address of image
deltax - impulse spacing in x direction
deltay - impulse spacing in y direction
nx - number of impulses in x direction
ny - number of impulses in y direction
xoff - the X offset for the first impulse from the upper
left corner of the image
yoff - the Y offset for the first impulse from the upper
left corner of the image
rows,cols - size of the image to be generated
The output image contains all zeros, except for the unit
impulses.
vimpulse returns a zero upon failure and a one upon suc-
cess.
SEE ALSO
vimpulse(1), intro(3), vipl(3), verror(3), vutils(3)
RESTRICTIONS
generates only byte data storage type output files.
AUTHOR
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.
*********************************************************************************
Documentation for avs module vmarr
INPUT
image pointer to VIFF structure containing image
data to be processed.
rows number of rows in kernel.
cols number of columns in kernel.
var the variance of the Gaussian
OUTPUT
image pointer to VIFF structure containing image
data after processing.
Return Value: 1 on success, 0 on failure.
DESCRIPTION
creates a convolution filter kernel using the following
equation. pixel = [-1/(PI*var*var)]*[1-r*r/(2*var)]*exp[-
r*r/(2*var)], where r is the calculated distance from the
center of the kernel.
A Marr filter is the Laplacian of a Gaussian, and is
intended for high lighting image edges and at the same time
smooth out the noise. vmarr creates image data of type
float.
SEE ALSO
vmarr(1), intro(3), vipl(3), verror(3), vutils(3) lvcus-
tom(3)
RESTRICTIONS
only generates float kernels, vconvert can change the image
to other data types.
AUTHOR
Richard Krukar, John Rasure, Robert Estes
COPYRIGHT
Copyright 1991, University of New Mexico. All rights
reserved.