Insight Toolkit Datasets - Visible Human Female

These datasets were created from the National Library of Medicine
Visible Human Female (VHF) physical crossections. There are two sets
of images for each dataset, one JPEG and one raw. The JPEG images can
be viewed with a variety of image viewers. The raw data has no
header. Each consecutive pixel contains a red, green and blue 8-bit
value. The VHF data was acquired at 1/3 mm pixel and 1/3 mm interslice
resolution. Each slice is stored in a single file. The files are named
as follows:

DATASET{jpg,raw}/PREFIX.SLICE#{a,b,c}.{jpg,raw.gz}

where DATASET is one of:
	CROP
	256N
	256S
	256E
and PREFIX is one of:
	CROP
	256
and SLICE runs from 1001 to 1299 and represents the millimeter
          location of the slice.
and a,b,c select the 0, 1/3 and 2/3 mm offset.

The jpg slices are not compressed and can be viewed as-is by
standard image viewers including browsers.

NOTE: The jpg slices are for reference only and should not be used
for analysis since the jpg compression scheme is lossey.

The raw slices have been compressed with gzip and must be decompressed.

Anatomic	CROPjpg	CROPraw

FirstRow      = 280;
FirstColumn   = 780;
NumberRows    = 670;
NumberColumns = 570;
_______________________

512	 512jpg	512raw

FirstRow      = 80;
FirstColumn   = 9;
NumberRows    = 512;
NumberColumns = 512;
______________________

256North 256Njpg 256Nraw

FirstRow      = 1;
FirstColumn   = 128;
NumberRows    = 256;
NumberColumns = 256;
____________________

256South 256Sjpg 256Sraw

FirstRow      = 255;
FirstColumn   = 128;
NumberRows    = 256;
NumberColumns = 256;
___________________

256East	256Ejpg 256Eraw

FirstRow      = 128;
FirstColumn   = 70;
NumberRows    = 256;
NumberColumns = 256;
___________________


Getting the Data
-----------------

The data is stored in a cvs repository. However, cvs has difficulty
handling large numbers of large files. Therefore, we must take care in
how we acquire that data from the cvs repoistory, since we can crash
the cvs server.

Please follow the subsequent steps to acquire a dataset.

1) cvs -d:pserver:your_name@public.kitware.com:/insight/cvsroot login
Enter your password
You only need to do this step once.

2) cvs -d:pserver:your_name@public.kitware.com:/insight/cvsroot co InsightData/VisibleWomanHead/README

3) Decide which dataset you want and checkout the script to get that
   dataset:
For example, if you want 256Njpg, then
   cvs -d:pserver:your_name@public.kitware.com:/insight/cvsroot co InsightData/VisibleWomanHead/256Njpg/GetData.bat

4) Now you should have a directory to hold the data. Go there and get
the individucal data files.
PLEASE DO NOT do a cvs update in this directory. Run the provided
script.
	cd InsightData/VisibleWomanHead/256Njpg

	./GetData.bat

This will take a long time to download. The good news is that if the
download is interrupted, you can run the GetData again and it will
resume where it ended.


------


The data set

       VisibleWomanSubsampledRGB.raw

is a subsampling of the RGB data taking every other pixels.
Its pixel spacing is then 2/3mm along each direction.
The subsampling was done by simply picking pixels, no smoothing
was done before the subsampling.

Pixel data is stored as  

               RGBRGBRGBRGBRGB.... 

using 1byte per component.

The file VisibleWomanSubsampledRGB.mhd is a MetaImage header
that allows to load the .raw file with ITK readers.