pcinfo Documentation

PCINFO - Calculate information needed for photoclinometry
Calculates and reports information useful for running the
programs pc2d and pcsi that perform two-dimensional photo-
clinometry (shape from shading, or topographic modeling based
on image brightness) using the finite-element method of Kirk
(1987), which is described further by Kirk et al. (2003).

One output, the "azimuth of characteristics," is the
direction in which topographic slopes (relative to a zero
surface or "datum" that may be inclined relative to the
camera) have a maximum effect on topographic shading in
the image.  Small slopes at right angles to this direction
have minimal effect on shading and hence are invisible.
The direction of the characteristics is generally toward
the sun, but also depends on the orientation of the datum
(determined from the image labels) and the photometric
properties of the surface.

The azimuth of characteristics is important because albedo
variations on the surface will cause artifacts in the
photoclinometric topography in the form of stripes along
this direction.  A small amount of striping in this direction
will also occur even if the albedo is completely uniform.
Standard post-processing of the output from pc2d or pcsi
therefore includes rotating the elevation model so that
the characteristic direction is aligned with the sample
axis, using the ISIS destriping filter procedure dstripe
to suppress stripes, and then re-rotating the model to its
original orientation.  The azimuth of characteristics is
returned to TAE for use in such processing, and is also
reported in the print file.

This program also calculates the memory needs of pc2d and
pcsi for a variety of options.  The memory required to use
the fast "SSIPSF-PI" method to generate an initial approxi-
mation to the topography depends not only on image size but
also on the azimuth of characteristics.  Memory required for
subsequent refinement of the model depends on which of the
available solution methods (successive over-relaxation,
incomplete Cholesky/conjugate gradient method, or direct
matrix factorization) are used and the number factors of
two by which the image resolution will be reduced during
processing to speed convergence of long spatial wavelengths
of the solution.  Memory requirements for all methods are
reported in the print file.

Programmer:  Randolph Kirk, U.S.G.S., Flagstaff

Input cube file name

NONE

Input subcube specifier

--

Type of datum model
(Zero-topography surface)
1=Plane 2=Sphere

1

Sample at which to calculate
geometry (rel to whole frame)

--

Line at which to calculate
geometry (rel to whole frame)

--

For Viking/Voyager, does image
still contain distortions?

NO

Radius of datum sphere in km
(If =< 0, label value used)

0.0

Center sample for sphere datum

--

Center line for sphere datum

--

Photometric function model
(LAMBER,LOMSEL,MIN,LUNLAM,
HAPHEN,HAPLEG,HAPH_S,HAPL_S)

LAMBER

Minnaert exponent for MIN

--

Lunar-Lambert weight for LUNLAM

--

Single-scattering albedo for
HAPHEN, HAPLEG, HAPH_S, HAPL_S

--

Opposition surge width for
HAPHEN, HAPLEG, HAPH_S, HAPL_S

--

Opposition surge strength for
HAPHEN, HAPLEG, HAPH_S, HAPL_S

--

Macroscopic surface roughness
in degrees, for HAPHEN, HAPLEG

--

Henyey-Greenstein asymmetry
parameter for single particle
phase function in HAPHEN, HAPH_S

--

2nd Henyey-Greenstein parameter
controls mix of +HG1, -HG1
components for HAPHEN, HAPH_S

--

Coefficient of P1(cos(phase))
in Legendre single particle
phase function for HAPLEG,
HAPL_S

--

Coefficient of P2(cos(phase))
in Legendre single particle
phase function for HAPLEG,
HAPL_S

--

ADDITIONAL NOTES:

Parm	Description
FROM	The image file upon which photoclinometry will be performed.
SFROM	Input subcube specifier for the FROM file. Selects the part of the input cube file upon which photoclinometry will be performed. The format for the SFROM parameter is ss-es(sinc):sl-el(linc):sb-eb(binc) where: ss = starting sample es = ending sample sinc = sample increment sl = starting line el = ending line linc = line increment sb = starting band eb = ending band binc = band increment The default value of NULL will select the entire input cube. If the line and sample increments are not equal, the map scale will not be accurate.
DATUMTYP	Type of model for the datum, or mean planetary surface. If DATUMTYP=1, the datum is a plane tangent at the image point at which geometric quantities are calculated (see X, Y). If DATUMTYP=2, the datum is a portion of a sphere. In this case the size and location of the sphere may be calculated from label information or overridden by user input (see RADIUS, XSPHERE, YSPHERE). Default is 1.
X	Affects the location in the image at which geometric quantities will be calculated from SPICE label information. These quantities will be calculated at sample X, line Y measured relative to the FULL camera frame from which the current data were extracted. If no value is given for X the sample coordinate of the center of the full frame will be used.
Y	Affects the location in the image at which geometric quantities will be calculated from SPICE label information. These quantities will be calculated at sample X, line Y measured relative to the FULL camera frame from which the current data were extracted. If no value is given for Y the line coordinate of the center of the full frame will be used.
DISTORTD	Do the images still contain camera distortions? If YES, the X,Y values from either the user inputs or the defaults for the center of the frame will be corrected before being used to find the illumination and viewing geometry. Default is NO (no correction).
RADIUS	Radius in km of the spherical datum model to be used if DATUMTYP=2 (q.v.). If RADIUS=<0, the planetary radius read from PLANET.SAV will be used, but if RADIUS>0 the value input will be used.
XSPHERE	Affects the center that will be used for a spherical datum model. If no value is given, the center of the disk will be calculated from the label information.
YSPHERE	Affects the center that will be used for a spherical datum model. If no value is given, the center of the disk will be calculated
PHOFUNC	This parameter selects the type of photometric function model used to describe the planetary surface. The parameters used differ between the photometric functions. PHOTOMETRIC FUNCTIONS TAE Full name Parameters ___ _________ __________ LAMBER Lambert none LOMSEL Lommel-Seeliger ("lunar") none LUNLAM Lunar-Lambert function L MIN Minnaert function K HAPHEN Hapke - Henyey-Greenstein WH,HG1,HG2, HH,B0,THETA HAPLEG Hapke - Legendre WH,BH,CH, HH,BH,THETA HAPH_S Hapke - Henyey-Gr. smooth WH,HG1,HG2 HAPL_S Hapke - Legendre smooth WH,BH,CH The functions are defined as follows, where phase is the phase angle, and u0 and u are the cosines of the incidence and emission angles, respectively Lambert FUNC=u0 Lommel-Seeliger FUNC=u0/(u0+u) Minnaert FUNC=u0**K * u**(K-1) Lunar-Lambert ("lunar" part is Lommel-Seeliger) FUNC=(1-L)*u0 + 2*L*u0/(u0+u) Hapke - Henyey-Greenstein Complete Hapke (1981; 1984; 1986) photometric model with Henyey-Greenstein single-particle phase function whose coefficients are HG1 and HG2, plus single scattering albedo WH, opposition surge parameters HH and B0, and macroscopic roughness THETA. Hapke - Legendre Similar to the previous except that the single particle phase function is a two-term Legendre polynomial with coefficients BH and CH. Hapke - Henyey-Greeenstein smooth Substantially simplified version of Hapke-Henyey-Greenstein function that omits the opposition effect as well as the (very slow) macroscopic roughness correction. For a smooth model with opposition effect, use the full Hapke-Henyey function with THETA=0. Hapke - Legendre smooth Simplified Hapke model with Legendre single particle phase function, no opposition surge, and no roughness correction. Hapke, B. W., 1981. Bidirectional reflectance spectroscopy 1: Theory. J. Geophys. Res., pp. 86,3039-3054. Hapke, B., 1984. Bidirectional reflectance spectroscopy 3: Corrections for macroscopic roughness. Icarus, 59, pp. 41-59. Hapke, B., 1986. Bidirectional reflectance spectroscopy 4: The extinction coefficient and the opposition effect. Icarus, 67, pp. 264-280. McEwen (1991) has compiled Hapke parameter estimates for many planets and satellites from a variety of sources. McEwen, A. S., 1991. Photometric functions for photo- clinometry and other applications. Icarus, 92, pp. 298-311.
K	Exponent that governs limb-darkening in the Minnaert photometric function: PHOFUNC=u0**K * u**(K-1). Values generally fall in the range from 0.5 ("lunar-like", almost no limb darkening) to 1.0 (Lambert function).
L	Weight that governs limb-darkening in the Lunar-Lambert photometric function: PHOFUNC=(1-L)*u0 + 2*L*u0/(u0+u). Values generally fall in the range from 0 (Lambert function) to 1 (Lommel-Seeliger or "lunar" function).
WH	Single-scattering albedo of surface particles, used if PHOFUNC=HAPHEN, HAPLEG, HAPH_S, or HAPL_S. See Hapke (1981).
B0	Magnitude of the opposition effect for the surface, used if PHOFUNC=HAPHEN, HAPLEG, HAPH_S, or HAPL_S. See Hapke (1984).
HH	Width parameter for the opposition effect for the surface, used if FUNC=HAPHEN, HAPLEG, HAPH_S, or HAPL_S. See Hapke (1984).
THETA	"Macroscopic roughness" of the surface as it affects the photometric behavior, used if PHOFUNC=HAPHEN or HAPLEG. This is the RMS slope at scales larger than the distance photons penetrate the surface but smaller than a pixel. See Hapke (1986). The roughness correction, which will be evaluated if THETA is given any value other than 0.0, is extremely slow.
HG1	Asymmetry parameter used in the Henyey-Greenstein model for the scattering phase function of single particles in the surface, used if PHOFUNC=HAPHEN or HAPH_S. See Hapke (1981). The two-parameter Henyey-Greenstein function is P(phase) = (1-HG2) * (1-HG1**2)/(1+HG1**2+2*HG1*COS(PHASE))**1.5 + HG2 * (1-HG1**2)/(1+HG1**2-2*HG1*COS(PHASE))**1.5
HG2	Second parameter of the two-parameter Henyey-Greenstein model for the scattering phase function of single particles in the surface, used if PHOFUNC=HAPHEN or HAPH_S. This parameter controls a the proportions in a linear mixture of ordinary Heneyey-Greenstein phase functions with asymmetry parameters equal to +HG1 and -HG1. See HG1 for the full formula.
BH	When PHOFUNC=HAPLEG or HAPL_S, a two-term Legendre polynomial is used for the scattering phase function of single particles in the surface P(phase) = 1 + BH * P1(COS(PHASE)) + CH * P2(COS(PHASE))
CH	When PHOFUNC=HAPLEG or HAPL_S, a two-term Legendre polynomial is used for the scattering phase function of single particles in the surface P(phase) = 1 + BH * P1(COS(PHASE)) + CH * P2(COS(PHASE))