Vicar labelled input files (IN,REF,OLDTP) IN - the image to be registered..byte format REF - the image used as reference..byte format OLDTP - optional file containing the raw tiepoints from a previous unfinished run of PICREG. See the format of NEWTP below.
Vicar labelled output files (NEWTP,PAR):
NEWTP - file containing the raw tiepoint values which is
updated for each tiepoint collected. Up to 10
records, each of 3200 bytes of real*4 data. The order of the
tiepoints is Lref,Sref,Lin,Sin.
The purpose of this file is to save tiepoints for
possible future runs of PICREG. This file is NOT
used as an interface to any other program!
WARNING: immediately upon startup, PICREG initializes
this dataset to contain 0 tiepoints. Therefore, the
user should be careful not to re-enter a previous NEWTP
file in this position, or the accumulated tiepoints will
be lost!
PAR - optional file to receive the formatted parameter file
for the chosen Geom program. It will be a Vicar2
parameter file in XVPOUT format.
Vicar SIZE field - ignored by PICREG default = (1,1,0,0)
Starting line of size field default = 1
Starting samp of size field default = 1
Number of lines of size field default = 0
Number of samps of size field default = 0
Interactive keyword which terminates the program. Note that the apostrophe can be ommitted. Both 'EXIT and EXIT are valid.
Interactive parameter which specifies, in response to a prompt, the Geom program which will be run on the output parameters. Valid = GEOMA, LGEOM, MGEOM. Default=GEOMA
Specifies that the display window for both images move up by the indicated number of pixels. Default=0.
Specifies that the display window for both images move down by the indicated number of pixels. Default=0.
Specifies that the display window for both images move left by the indicated number of pixels. Default=0.
Specifies that the display window for both images move right by the indicated number of pixels. Default=0.
specifies that the display window for the left image move up by the indicated number of pixels. default=0.
specifies that the display window for the right image move up by the indicated number of pixels. default=0.
specifies that the display window for the left image move down by the indicated number of pixels. default=0.
specifies that the display window for the right image move down by the indicated number of pixels. default=0.
specifies that the display window for the left image move left by the indicated number of pixels. default=0.
specifies that the display window for the right image move left by the indicated number of pixels. default=0.
specifies that the display window for the left image move right by the indicated number of pixels. default=0.
specifies that the display window for the right image move right by the indicated number of pixels. default=0.
specifies that the display window for the left image begin at the indicated pixel. default=1.
specifies that the display window for the right image begin at the indicated pixel. default=1.
specifies that the display window for the left image begin at the indicated pixel. default=1.
specifies that the display window for the right image begin at the indicated pixel. default=1.
specifies that both display windows are to begin at (1,1).
will move both windows so that the tiepoint indicated is located in the center of each window. This is true unless the tiepoint is located near the left edge, or near the top edge of the image. In this case, the image will be displayed starting at sample 1, or line 1, and the tiepoint will be at the corresponding sample or line. default=1
Height (in pixels) of tiepoint numbers on screen. The program starts with a value of 18 if the screen size is bigger than 512 by 512 or a value of 12 if the screen size does not exceed 512 by 512. 1<= textsize<=99. The user can effectively turn off the numbering by setting TEXTSIZE to 1 as long as desired.
Indicates an integer which specifies the magnification or reduction applied to both images. If the value is >0, the pic is entlarged by that factor. If it is <0, the pic is reduced by the factor 1/value. Compressions can facilitate fewer window movements, give a better idea of where the user is in the file , and display the tiepoint in a more global sense. zoom=0 is interpreted as no zoom, the same as zoom=1. The maximum value for zoom is 8. There is no minimum value for zoom other than the limit of -2**31 for valid integers. This means the compression may be as large as is useful. The default for zoom is 1.
the left image will be zoomed by this factor. default=1
the right image will be zoomed by this factor. default=1
stretch=(lo,hi) specifies that both images will have a linear contrast stretch applied. lo goes to 0 dn, hi goes to 255 dn. default=(0,255). (There is no restriction keeping hi from being greater than 255, but this is generally not done since it makes the image display darker. It would not allow the image files to contain DNs outside the range 0 to 255 since the images must have byte format.) Currently there is no provision for stretching the left and right sides independently. If the two input images require separate stretches, the stretches must be performed prior to entering PICREG.
linear=(lo,hi) specifies that both images will have a linear contrast stretch applied. lo goes to 0 dn, hi goes to 255 dn. default=(0,255). (There is no restriction keeping hi from being greater than 255, but this is generally not done since it makes the image display darker. It would not allow the image files to contain DNs outside the range 0 to 255 since the images must have byte format.) Currently there is no provision for stretching the left and right sides independently. If the two input images require separate stretches, the stretches must be performed prior to entering PICREG.
will cause the sl, ss, and zoom factor to be printed for each image.
valid=(corr,point) determines whether PICREG will accept the user cursor locations as the final tiepoint values (point) or try to optimize the locations using a fft cross-correlation technique. default=point. If MODE=CORR is selected, POWER defaults (is set) to 6.
Tells PICREG to use fft correlation. The value indicates the the size of the correlation window. size= 2**power. The default is 0, which means to not use fft correlation, unless MODE=CORR (or 'CORR) is selected. If MODE=CORR is selected, POWER defaults (is set) to 6. Thus, when fft correlation is selected, the correlation window size is 64 by 64 (pixels) unless explicitly changed using the POWER parameter.
VALID=(HPF,NHPF) determines whether, in the fft correlation, the vertical and horizontal frequencies which lie on axes passing through the DC term are zeroed out (hpf) or not (nohpf). default=hpf.
This keyword determines whether or not phase correlation is used when correlating tiepoints. In phase correlation, the complex fourier transform coefficients are divided by their amplitudes before computing the correlation function.
valid=(inte,noin) determines whether, in the fft correlation, the centroid of the cor- relation peak is determined to nearest-pixel (noin) or sub-pixel (inte) accuracy. If 'inte is specified, PICREG tries to repeat the correlation until it converges to within .1 pixels. If 'noin is specified, PICREG tries to repeat the correlation until it converges to within 1.1 pixels. default=noin. (See also Example 2 in the main HELP for PICREG.)
invokes the area mode of correlation and supplies the size of the box used. In this mode, PICREG will compute two buffers containing the sums of rows and of columns of dn's in an area about the tiepoint location. The final positions will be set to the minimum or maximum positions of each buffer. default=20.
valid=(min,max) indicates whether, in the area mode, to pick the maximum or the minimum positions of the buffers. default=min.
specifies which tiepoint number is to be deleted and redone. default=0
Specifies a range of tiepoint numbers to be deleted. default=(0,0). Note that any following tiepoints will be renumbered.
results in a display on the left image graphics plane of tiepoint displacements from their location in the left image relative to their location in the right image. This is done relative to the first tiepoint ( it will always show no displacement) and is only done for those tiepoint presently displayed.
picks a type of least squares surface, and fits it to the current set of tiepoints. The meanings of the values are: 1. rotation + offset 2. rotation + offset + scale 3. first order unconstrained 4. scale + offset 5. offset 6. second order unconstrained 7. third order unconstrained PICREG will print the mean error and the number and error of the worst fitting point. This keyword is used to check for tiepoint location errors and to test how well a specifc surface fits prior to using that surface to generate an output grid of regularly spaced tiepoints for use with a Geom program. 1<= fit <= 7. default = 1
This keyword formats the second output file for later use with TIECONM, GEOMA, LGEOM, or MGEOM. Value values and their meanings are: 1-7. The corresponding surface as described under the fit parameter (also below) is used to generate a nhor by nver grid of regularly spaced tiepoints in GEOMA, LGEOM, or MGEOM format. These tiepoints, along with all necessary parameters are then written to the second output file. The meanings of the TPFORM values in the range 1 to 7 are: 1. rotation + offset 2. rotation + offset + scale 3. first order unconstrained 4. scale + offset 5. offset 6. second order unconstrained 7. third order unconstrained 8. The region of the reference picture which is encompassed by the output tiepoint network will be samlpled in a nhor by nver grid. At each grid intersection, a transformation will be computed by least squares using the relations: y1 = ax + by + c x1 = dx + ey + f between the reference pic (x,y) and the input picture (x1,y1) using the n closest tiepoints to that grid location. (n is set using the USE parameter.) 9. This invokes the TIECONM option. No re-sampling is done in this case. Instead, the raw tiepoints and all necessary TIECONM parameters are written to the second output. This parameter file can then be used with TIECONM which will in turn format the tiepoints for use with GEOMA, LGEOM or MGEOM. The TIECONM triangulation method for surface fitting is useful in cases where an image has areas of significant distortion. In any case, the user will be prompted for which Geom program he plans to use. If GEOMA is specified, the output grid will be transformed into triangles. Applicable keywords (nhor,nver,minl,mins,maxl,maxs,use) should be specified prior to or on the same line as the tpform keyword.
Specifies the number of columns of tiepoints in the output grid, for the case of TPFORM = 6, 7, or 8. This is the number of tiepoints in a horizontal line of the output tiepoint grid. If TPFORM <= 5 is chosen, then this parameter is ignored, and NHOR=2 is used (since 4 points suffice to determine a first-order transformation). Note that if GEOMA format is chosen for the tiepoints, then the value of NHOR is almost doubled (NHOR = 2*NHOR-1) in order to make triangular tiepoint areas. Default=20.
Specifies the number of rows of tiepoints in the output grid, for the case of TPFORM = 6, 7, or 8. This is the number of tiepoints in a vertical line of the output tiepoint grid. If TPFORM <= 5 is chosen, then this parameter is ignored, and NVER=2 is used (since 4 points suffice to determine a first-order transformation). Default=20.
minimum line number of the output grid tiepoints. default is based on raw tiepoints.
minimum samp number of the output grid tiepoints. default is based on raw tiepoints.
maximum line number of the output grid tiepoints. default is based on raw tiepoints.
maximum samp number of the output grid tiepoints. default is based on raw tiepoints.
valid=(print,noprint) enables the printing of least squares fit information when the fit is specified and results in a printout of all raw tiepoint locations upon exit default=noprint.
the number of tiepoints used to determine each output tiepoint location. Applies to tpform=8 only. 4<= use <= 10 . default=4
selects the shape of the cursor by specifying its number. The number
scheme is installation or display-type dependent, and may be changing.
At present, the following are available on the DeAnza:
. | -----
1 = . . . 2 = ---|--- 3 = | . |
. | |_____|
______
|| \ / | |
4 = == == 5 = complement 6 = \/ 7 = |------|
|| of 4 /\ |______|
/ \
The cursor shapes for values 1 to 6 are also valid under the X Window System.
POSITION values are used instead of the trackball cursor position. POSITION values are in terms of image line and sample coordinates. This parameter is useful in testing PICREG since it bypasses the trackball and allows testing to be repeatable.