COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\gdef\pagstrt{F}
C00008 ENDMK
C⊗;
\gdef\pagstrt{F}
\def\fontsize{\tenpoint}
\fontsize
\baselineskip 13pt
\parskip 10pt plus 3 pt minus 3 pt
\parindent .7vu
\head{Progress Report}
\vspssml
\ctrline{Period ending February 28, 1982}
\vsplrg

Theoretical analysis  has  led to  solutions  for the  synthesis  of  high
resolution  cursor  patterns.   The   algorithms  which  implement   these
solutions have been  outlined.  Solutions  have been found  for the  cases
proposed: dot and straight line.  Synthesis of 3d cursors from 2d  cursors
has been done  for the  Stanford stereo station  for dot  cursors and  for
orthogonal vector  triples for  measuring  surface normals.   The  methods
under development  could  also  be  used for  those  triples  for  surface
normals.

Theoretical solutions and outlines of algorithms for their  implementation
have been developed for two  additional problems of potential interest  to
DMA.  These tasks were not included in the proposal.  One is the synthesis
of  high  resolution  graphics  composed  of  circular  arcs.   Reasonable
graphics curves can  be synthesized by  splines with a  basis of  circular
arcs, i.e.   piecewise  smooth  curves  composed  of  circular  arcs  with
continuous tangent vector.

The second task for which a solution  has been obtained but which was  not
included in the  proposal is the  problem of scrolling  an image  smoothly
past a  fixed cursor.   This is  current  practice in  using film  and  is
convenient for coarse location.  This  method suggests a very  inexpensive
modification to hardware of displays to enable smooth scrolling.

In the last few years, several  papers have appeared in the literature  on
this subject.  We have collected and read relevant papers.

We have  identified  several  tests  of  positional  resolution  for  high
resolution cursors.  We are enlisting members of the perceptual psychology
faculty of Stanford to aid in evaluation and in providing a background  in
psychophysics results relevant to this effort.  We have available from SRI
several image  sequences of  Fort Belvoir.   We do  not know  yet  whether
ground truth is available  for these images,  especially for the  cultural
sites.

Implementation has encountered several  difficulties.  However, we  intend
to succeed  in  spite of  them.   Our  Grinnell display  has  had  several
problems which make it  completely unusable.  It is  at the factory  being
repaired.  We do not yet have a date for its return.  They have been  very
slow in the past.  If the Grinnell is not available in time, we will  find
other ways of display, for example, using the SRI Testbed if possible.

Also, the student who is doing the  program has been diverted to handle  a
major emergency, and has  been replaced.  Our  Datadisc system has  nearly
died; there are few usable  channels left.  Consequently, it is  difficult
to get  terminals on  the SAIL  system.  The  student is  now finishing  a
replacement for the Datadisc terminal system.   He may be finished in  two
weeks,  but  it  is  my  experience  that  these  things  drag  out.   His
replacement is capable, and I think will succeed according to our  revised
schedule.

\ljust{Here's the schedule:}
\vspsml
{\halign{#\hfil⊗\hfil#\hfil\cr
gray scale dot cursor:⊗March 21\cr
3d gray scale dot cursor:⊗April 7\cr
theoretical analysis of⊗\cr
\hfil accuracy:⊗April 21\cr
gray scale lines:⊗April 28\cr
reports:⊗monthly from Feb;\cr
⊗final report\cr
perceptual evaluation⊗\cr
\hfil of 2d image resolution:⊗April 28\cr
analysis of 3d resolution:⊗May 7\cr
perceptual evaluation⊗\cr
\hfil of 3d results:⊗\cr
}}