
		 The EyeBox package and example

1.  Introduction 

	This README file describes the EyeBox computational tool
for analyzing landmark location uncertainty.  The tool is
described in Chapter 6 of the book "Precision Landmark Location
for Machine Vision and Photogrammetry," published by Springer
(London, 2008).

	The EyeBox computational tool can determine the
Cramer-Rao bound on landmark location uncertainty, and perform a
Monte-Carlo study, investigating landmark location uncertainty.

	The EyeBox computational tool is driven from a text
configuration file. This distribution is provided with an example
configuration file (file Pulnix35mm-Model.txt) that runs the
nominal case described in chapters 6, 7 and 8 of "Precision"

2.  Disclaimer


	This software has been developed as an analysis package
as part of a PhD study.  It has not received the attention that
is needed to make it a friendly, easy to use program.  As such,

			  USER BEWARE

the documentation is limited (description is a fairer term) and
the user interface is idiosyncratic.  This README is intended
only to enable the user to run the example case.  From that
point, adaptation of the system to other applications is entirely
up to the user

	This software is freely available and comes with no
warranty.  This program is distributed in the hope that it will
be useful, but WITHOUT ANY WARRANTY, to the extent permitted by
law; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.


3.  Configuration of EyeBox 

	Routine EyeConfig.m reads a text file containing
configuration information.  Example program EyeBoxExample.m sets
a pointer to "Pulnix35mm-Model.txt" and calls EyeEntity.m to
execute the example.

	The configuration file contains information about the
Camera characteristics, Landmark attributes and experimental
setup details. To update the configuration information, open the
file with a tool that formats in a user friendly manner the
presentation of a tab-separated file e.g.  Excel. Make changes to
the relevant rows and save back in tab-separated text format.

4.  Running the example case 

	An example case is included.  To run the EyeBox example
case:

  a) Unpack file EYESystem.zip

  b) Modify the configuration file Pulnix35mm-Model.txt as
     desired

  c) Launch matlab with EyeBox as the working directory

  d) In matlab, execute >> EyeBoxExample

	Because of the many numerical integrations,
EyeBoxExample.m requires about 1.5 hours to run on an Intel Core
2 Duo.  


5.  What is produced by EyeBoxExample

AnalysisPhase1: 

	This directory contains MATLAB data files with
the output generated by EyeEntity consisting on a high resolution
numerical model of the optical flux function. In addition, these
files contain camera and landmark configuration information that
are required in the next processing steps.

AnalysisPhase2a: 

	This directory contains MATLAB data files corresponding
to the Raster Spatial Sampling executed over the files located in
AnalysisPhase1.


AnalysisPhase3a: 

	This directory contains MATLAB data files generated after
the amplitude quantization of the Raster Spatial Sampling located
in AnalysisPhase2a.



AnalysisPhase4a: 

	This directory contains MATLAB data files with the
results of the set of estimators used over each digital image
located in AnalysisPhase3a.


AnalysisPhase5a: 

	This directory contains MATLAB data files containing
information about the confidence intervals generated with the
estimates stored in AnalysisPhase4a.

Figures:  

	A figure similar to that shown in
EyeBoxExampleOutputPlot.pdf should be produced.

For more information, refer to the functional architecture
pictorial shown in file 

    EyeBox_Functional_Architecture.ppt  

and to "Precision Landmark Location."

Cheers, 

Jose Gutierrez,  joe.natacha@ieee.org
Brian Armstrong, bsra@uwm.edu

