MicroCol24-example:
-------------------
This is an example of phytoplankton pictures taken under the microscope with
a digital camera (jpeg files in 24bit color with no extra precaution like
background substration or color calibration; only the size of a pixel is
determined and provided in the metadata, see MicroCol24-example.txt/xls files,
'PixelSize' column). There images are provided by Elisabeth Nezan (IFREMER,
France). They are typical examples of phytoplankton microphotographies.

Segmentation of the image is done in three steps: (1) simple threshold
relatively high above the background to retrieve dark parts, (2) threshold based
on contrasted areas using a Sobel filter, and (3) a threshold closer to the
background after smoothing with the Rolling ball algorithm. All three masks are
combined together ultimately. It is also a demonstration of the use of an
"ImportTemplate.zie" file together with a table describing variable metadata
fields from image to image.

To run this example:
--------------------
- Download and unzip "MicroCol24-example.zip" in your 'ZooPhytoImage Examples'
  directory, or anywhere you like to place it.
  
- Start Zoo/PhytoImage and import the data (second button, select 'Table and
  ImportTemplate.zie' file type at the bottom of the dialog box and then,
  select 'MicroCol24-example.txt' file). The import mechanism compiles the .zim
  files (metadata) for each sample and renames the images accordingly (in the
  '_work' subdir).
  
- Once files are ready, they can be analyzed (third button, switch to ImageJ).
  In ImageJ, select (Plugins -> ZooPhytoImage -> Microscope Color). Select one
  of the .zim files you just created and watch the analysis live on screen. You
  can examine the mask and other temporary files in the '_work' subdirectory.
  
- Once the analysis is done, you can close ImageJ and switch back to the
  Zoo/PhytoImage assistant. Click on the fourth button to finalize the two .zid
  files corresponding to this analysis. You should end up with two .zid files
  (Zoo/PhytoImage Data) plus all the original files moved to the '_raw' subdirectory.
  
What next...:
-------------
The .zid files are key files in Zoo/PhytoImage. They contain everything you need
for manual and automatic identification of the particules and for the calculation
of summary statistics (abundances, biomasses, size spectra, etc.). At this point,
you can archive (on DVD, external hard disks, etc.) the original data that are
now in the '_raw' subdirectory to free space on your machine, and safely work
with the .zid files only.

With only two samples and 5 images in each, you cannot do much more, but you
are supposed to collect more data, then to train a classifier, check its
performances and finally use it to identify particles in your samples.
For example with more data, switch to the examples whose name ends with
'train&data'. We have used a set of a couple of thousands of such images of
phytoplancton all around French coast, and we were able to separate about 20
taxa with an overall accuracy close to 80% with Zoo/PhytoImage. We were also
able to differenciate among 6 species of Dynophysis with similar performances.