Skip to content

Commit

Permalink
Finished Editing Intro.
Browse files Browse the repository at this point in the history
  • Loading branch information
Ross Beyer authored and Ross Beyer committed Oct 23, 2009
1 parent ee90a39 commit 643a56d
Show file tree
Hide file tree
Showing 3 changed files with 81 additions and 74 deletions.
8 changes: 7 additions & 1 deletion docs/book/asp_book.tex
Expand Up @@ -4,7 +4,8 @@
% \usepackage{moreverb}
% \usepackage{float}
\usepackage{subfigure}
\usepackage{graphicx, amsfonts, psfrag, fancyhdr, layout, appendix, subfigure}
\usepackage{graphicx}
\usepackage{amsfonts, psfrag, fancyhdr, layout, appendix, subfigure}
\usepackage{wrapfig}
\usepackage{color}
\definecolor{violet}{rgb}{0.5,0,1}
Expand Down Expand Up @@ -102,7 +103,12 @@
\acro{NASA}{National Aeronautics and Space Administration}
\acro{DEM}{digital elevation model}
\acro{MGS}{Mars Global Surveyor}
\acro{USGS}{United States Geological Survey}
\acro{ISIS}{Integrated Software for Imagers and Spectrometers}
\acro{MER}{Mars Exploration Rover}
\acro{MRO}{Mars Reconnaissance Orbiter}
\acro{LRO}{Lunar Reconnaissance Orbiter}
\acro{MOC}{Mars Orbiter Camera}
\end{acronym}


Expand Down
4 changes: 2 additions & 2 deletions docs/book/bibliography.bib
Expand Up @@ -262,7 +262,7 @@ @INPROCEEDINGS{1997LPI....28..387G
@unpublished{
ISIS_website,
author = {U.S. Geological Survey, Flagstaff, AZ},
title = {Integrated Software for Imagers and Spectrometers (ISIS)},
title = {Integrated Software for Imagers and Spectrometers ({ISIS})},
url = {http://isis.astrogeology.usgs.gov/},
year= {2009}
}
Expand Down Expand Up @@ -599,4 +599,4 @@ @inproceedings{broxton:isvc09
author={ Michael Broxton and Ara V. Nefian and Zachary Moratto and Taemin Kim and Michael Lundy and Aleksandr V. Segal },
booktitle = {{to appear in the Proceedings of the 5th International Symposium on Visual Computing}},
year = "2009"
}
}
143 changes: 72 additions & 71 deletions docs/book/introduction.tex
@@ -1,40 +1,40 @@
\chapter{Introduction}

The NASA Ames Stereo Pipeline (ASP) is a suite of automated geodesy \&
\acresetall

The \acsu{NASA} \ac{ASP} is a suite of automated geodesy and
stereogrammetry tools designed for processing planetary imagery
captured from orbiting and landed robotic explorers on other planets.
It was designed to process stereo imagery captured by NASA spacecraft
and produce cartographic products including digital elevation models
(DEMs), ortho-projected imagery, and 3D models. These data products
are suitable for science analysis, mission planning, and public
outreach.
It was designed to process stereo imagery captured by \ac{NASA}
spacecraft and produce cartographic products including \acp{DEM},
ortho-projected imagery, and 3D models. These data products are
suitable for science analysis, mission planning, and public outreach.

\begin{figure}[tb]
\centering
\includegraphics[width=6.5in]{images/introduction/p19view2.png}
\caption{This 3D model was generated from Mars Orbiter Camra image
pair M01-00115 and E02-01461 (34.66N, 141.29E). The complete
\caption{This 3D model was generated from \ac{MOC} image
pair M01/00115 and E02/01461 (34.66N, 141.29E). The complete
stereo reconstruction process takes approximately five minutes on
a 3.0GHz workstation for input images of this size (1024x8064
pixels). This model, shown here without vertical elevation
exaggeration, is roughly 2-km wide in the cross-track
a 3.0~GHz workstation for input images of this size ($1024 \times 8064$
pixels). This model, shown here without vertical
exaggeration, is roughly 2~km wide in the cross-track
dimension. }
\label{fig:p19}
\end{figure}

\section{Background}

The Intelligent Robotics Group (IRG) at the NASA Ames Research Center
has been developing 3D surface reconstruction and visualization
capabilities for planetary exploration for more than a decade. First
demonstrated during the Mars Pathfinder Mission, the IRG has delivered
tools providing these capabilities to the science operations teams of
the Mars Polar Lander mission, the Mars Exploration Rover mission, the
Mars Reconaissance Orbiter mission, and most recently the Lunar
Reconaissance Orbiter Mission. A critical component technology
enabling this work is the Ames Stereo Pipeline. The Stereo Pipeline
generates high quality, dense, texture-mapped 3D surface models from
stereo image pairs.
The \ac{IRG} at the NASA Ames Research Center has been developing
3D surface reconstruction and visualization capabilities for planetary
exploration for more than a decade. First demonstrated during the
Mars Pathfinder Mission, the \ac{IRG} has delivered tools providing
these capabilities to the science operations teams of the Mars Polar
Lander (MPL) mission, the \ac{MER} mission, the \ac{MRO} mission,
and most recently the \ac{LRO} mission. A critical component
technology enabling this work is the \acf{ASP}. The Stereo Pipeline
generates high quality, dense, texture-mapped 3D surface models
from stereo image pairs.

Although initially developed for ground control and scientific
visualization applications, the Stereo Pipeline has evolved in recent
Expand All @@ -44,42 +44,43 @@ \section{Background}
topography is often derived from stereo pairs captured from orbit.
Orbital mapping satellites are sent as precursors to planetary bodies
in advance of landers and rovers. They return a wealth of imagery and
other data that helps mission planners and scientist identify areas
other data that helps mission planners and scientists identify areas
worthy of more detailed study. Topographic information often plays a
central role in this planning and analysis process.

Our recent development of the Stereo Pipeline coincides with a period
of time when NASA orbital mapping missions are returning orders of
magnitude more data than ever before. Data volumes from the Mars and
Lunar Reconaissance Orbiter missions now mesaure in the tens of
Terabytes. There is growing consensus that Existing processing
techniques, which are still extremely human intensive and expensive,
are no longer adequate to address the data processing needs of NASA
and the Planetary Science community. To pick an example of particular
relevance, the HiRISE camera website lists 1353 stereo pairs at the
time of this writing \cite{HiRISE_website}. Of these, only a few tens
of stereo pairs have been processed to date; mostly on human-operated,
high-end photogrammetric workstations. It is clear that much more
value could be extracted from this valuable raw data if a more
streamline, efficient process could be developed.
Our recent development of the Stereo Pipeline coincides with a
period of time when \ac{NASA} orbital mapping missions are returning
orders of magnitude more data than ever before. Data volumes from
the Mars and Lunar Reconnaissance Orbiter missions now measure in
the tens of Terabytes. There is growing consensus that existing
processing techniques, which are still extremely human intensive
and expensive, are no longer adequate to address the data processing
needs of \ac{NASA} and the Planetary Science community. To pick
an example of particular relevance, the HiRISE camera Web site lists
1353 stereo pairs at the time of this writing \citep{HiRISE_website}.
Of these, only a few tens of stereo pairs have been processed to
date; mostly on human-operated, high-end photogrammetric workstations.
It is clear that much more value could be extracted from this
valuable raw data if a more streamlined, efficient process could be
developed.

The Stereo Pipeline was designed to address this very need. By
applying recent advances in robotics and computer vision, we have
created an {\em automated} process that is capable of generating high
quality DEMs with minimal human intervention. Users of the Stereo
quality \acp{DEM} with minimal human intervention. Users of the Stereo
Pipeline can expect to spend some time picking a handful of settings
when they first start processing a new type of imagery, but once this
is done the Stereo Pipeline can be used to process tens, hundreds, or
even thousands of stereo pairs without further adjustment. With the
release of this software, we hope to encourage the adoption of this
toolchain at institutions that run and support these remote sensing
tool chain at institutions that run and support these remote sensing
missions. Over time, we hope to see this tool incorporated into
ground data processing systems alongside other automated image
processing pipelines. As this tool continues to mature, we believe
that it will be capable of producing digital elevation models of
exceptional quality without any human intervention.

\section{Human vs. Computer : When to Choose Automation}
\section{Human vs. Computer: When to Choose Automation}

When is it appropriate to choose automated stereo mapping over the use
of a conventional, human-operated photogrammetric workstation? This
Expand All @@ -91,23 +92,23 @@ \section{Human vs. Computer : When to Choose Automation}
where human lives or considerable capital resources are at risk. In
particular, maps for landing site analysis and precision landing
absolutely require the benefit of an expert human operator to
eliminate obvious errors in the DEM; and also to guarantee that the
eliminate obvious errors in the \ac{DEM}; and also to guarantee that the
proper procedures have been followed to correct satellite telemetry
errors so that the data have the best possible geodetic control.

When it comes to using DEMs for scientific analysis, both techniques
have their merits. Human-guided stereo reconstruction produces DEMs
When it comes to using \acp{DEM} for scientific analysis, both techniques
have their merits. Human-guided stereo reconstruction produces \acp{DEM}
of unparalleled quality that benefit from the intuition and experience
of an expert. The process of building and validating these DEMs is
of an expert. The process of building and validating these \acp{DEM} is
well established and accepted in the scientific community.

However, only a limited number of DEMs can be processed to this level
However, only a limited number of \acp{DEM} can be processed to this level
of quality. For the rest, automated stereo processing can be used to
produce DEMs at a fraction of the cost. The results are not
produce \acp{DEM} at a fraction of the cost. The results are not
necessarily less accurate than those produced by the human operator,
but they will not benefit from the same level of scrutiny and quality
control. As such, users of these DEMs need to be able to identify
potential issues, and to be on the lookout for errors that may result
control. As such, users of these \acp{DEM} must be able to identify
potential issues, and be on the lookout for errors that may result
from the improper use of these tools.

We recommend that all users of the Stereo Pipeline take the time to
Expand All @@ -119,26 +120,26 @@ \section{Human vs. Computer : When to Choose Automation}
\section{The USGS Integrated Software for Imagers and Spectrometers}

This version of the Stereo Pipeline must be installed alongside a
compatible version of the USGS Integrated Software for Imagers and
Spectrometers (ISIS). ISIS is widely used in the planetary science
community for processing raw spacecraft imagery into high level data
products of scientific interest such as map projected and mosaicked
imagery \cite{2004LPI....35.2039A, 1997LPI....28..387G, ISIS_website}.
We chose ISIS because (1) it is widely adopted by the planetary
science community, (2) it contains the athoritative collection of
geometric camera models for planetary remote sensing instruments, and
(3) is open source software that is easy to leverage.
compatible version of the \ac{USGS} \ac{ISIS}. \ac{ISIS} is widely
used in the planetary science community for processing raw spacecraft
imagery into high level data products of scientific interest such
as map projected and mosaicked imagery \cite{2004LPI....35.2039A,
1997LPI....28..387G, ISIS_website}. We chose \ac{ISIS} because (1)
it is widely adopted by the planetary science community, (2) it
contains the authoritative collection of geometric camera models
for planetary remote sensing instruments, and (3) it is open source
software that is easy to leverage.

By installing the Stereo Pipeline, you will be adding an advanced
stereo image processing capability that can be used in your existing
ISIS workflow. The Stereo Pipeline supports the ISIS ``cube'' file
fromat, and can make use of the ISIS camera models and ancillary
information (i.e. SPICE kernels) for imagers on many NASA spacecraft.
The use of this single standardized set of camera models ensures
consistency between products generated in the Stereo Pipeline and
those generated by ISIS. Also by leveraging ISIS camera models, the
Stereo Pipeline can process stereo pairs captured by just about any
NASA mission.
\ac{ISIS} work flow. The Stereo Pipeline supports the \ac{ISIS}
``cube'' (\texttt{.cub}) file format, and can make use of the \ac{ISIS}
camera models and ancillary information (i.e. SPICE kernels) for
imagers on many \ac{NASA} spacecraft. The use of this single standardized
set of camera models ensures consistency between products generated
in the Stereo Pipeline and those generated by \ac{ISIS}. Also by
leveraging \ac{ISIS} camera models, the Stereo Pipeline can process
stereo pairs captured by just about any \ac{NASA} mission.

As an additional note, the Stereo Pipeline can also process arbitrary,
non-ISIS images with accompanying camera information, but doing so
Expand All @@ -156,25 +157,25 @@ \section{The USGS Integrated Software for Imagers and Spectrometers}
%
%CORRELATION
%
%BUNDLE ADQJUSTMENT
%BUNDLE ADJUSTMENT


\section{Getting Help}

All bugs, feature requests, and general discussion should be sent to
the Ames Stereo Pipeline user mailing list:
\begin{quote}
\indent \url{stereo-pipeline@lists.nasa.gov}
\indent \href{mailto:stereo-pipeline@lists.nasa.gov}{stereo-pipeline@lists.nasa.gov}
\end{quote}
To subscribe to this list, send an empty email message with the
subject 'subscribe' (without the quotes) to:
subject `subscribe' (without the quotes) to:
\begin{quote}
\indent \url{stereo-pipeline-request.lists.nasa.gov}
\indent \href{mailto:stereo-pipeline-request@lists.nasa.gov}{stereo-pipeline-request@lists.nasa.gov}
\end{quote}
To contact the lead developers and project manager directly, send mail
to:
\begin{quote}
\indent \url{stereo-pipeline-owner@lists.nasa.gov}
\indent \href{mailto:stereo-pipeline-owner@lists.nasa.gov}{stereo-pipeline-owner@lists.nasa.gov}
\end{quote}

\section{Warnings to users of the Ames Stereo Pipeline ALPHA}
Expand All @@ -190,7 +191,7 @@ \section{Warnings to users of the Ames Stereo Pipeline ALPHA}
robust, they have not been systematically tested or rigorously
compared to other methods in the peer-reviewed literature. We have a
number of efforts underway to carefully compare Stereo
Pipeline-generated data products to those produced uisng established
Pipeline-generated data products to those produced using established
processes, and we will publish those results as they become available.
In the meantime, {\bf we strongly recommend that you consult us first
before publishing any results based on the cartographic products
Expand Down

0 comments on commit 643a56d

Please sign in to comment.