Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: imageNCIF

 
> 	I would like to comment on the question of binary or ascii.  One 
> of the basic rules of cifs is that they must be in ascii and it is not 
> easy at the moment to see how one could arrange an escape in the middle 
> of a file to allow the inclusion of binary.  From the point of view of 
> comcifs, the use of binary files is definitely a problem and we would 
> like to see ascii used throughout.  There may be constructive ways of 
> dealing with the problem, but the simplest would be to keep everything in 
> ascii.
> 
It's my turn

ascii for images seem difficult to accept. 
Any pixel is stored on a minimum of one byte, which means it cannot be handled with one Ascii character only. Thus we would need two characters. Always keeping in mind that we only use Ascii, compression will be very poor and we can expect images files to be at least 1.5 and up to twice larger when written with ASCII than when using binary uncompressed format. 
This seems hard to accept and I am very pessimistic about the success of an image format if we cannot solve this point. 

Thus I think we should stat to discuss this point first.

Following is my answer about the other points mentioned by Andy.
My experience arises from X-Ray topography where we store TV images and from simulations of the same images. We have written an image treatment software
(you can consult the notice on the server:
http://www.lmcp.jussieu.fr/sincris/ in the software section. Its name is triton) and we had to introduce our own image format. Thus I already had some thought about headers.

We need a header, written in ASCII and it should be easy, in the future, to add new fields. CIF allows this easily. Among the information that we use:

- a comment giving an explanation about the image

- a reference: each scientist references images his/her own way

- a date  when image was recorded / created.

- size: - 1 - number of lines and columns
        - 2 - field of view or dimensions of pixels

- type: integer, floating point number, complex together with lenght of word, i.e. 1, 2, 4 for integers, 4, 8 for reals as well as for complex. (We use complex to store Fourier transforms

- statistics about the image: minimum, maximum, mean value as well as standart deviation

This is just a begining.

Best regards,

Yves



       -------------------------------------------------------------
       :    Yves EPELBOIN                                          :
       +-----------------------------------------------------------+
       :    Lab. Mineralogie-Cristallographie, UA 09 CNRS,         :
       :    Case 115, 4 place Jussieu                              :
       :    F-75252 PARIS CEDEX 05                                 :
       :    phone: 33 1 44 27 52 11                                :
       :    Fax: 33 1 44 27 37 85                                  :
       :    E-Mail: epelboin@lmcp.jussieu.fr                       :
       :    WWW   : http://www.lmcp.jussieu.fr/~epelboin           :
       +-----------------------------------------------------------+
       : UNIVERSITE PIERRE ET MARIE CURIE        PARIS     FRANCE  :
       +-----------------------------------------------------------+



Reply to: [list | sender only]
International Union of Crystallography

Scientific Union Member of the International Science Council (admitted 1947). Member of CODATA, the ISC Committee on Data. Partner with UNESCO, the United Nations Educational, Scientific and Cultural Organization in the International Year of Crystallography 2014.

International Science Council Scientific Freedom Policy

The IUCr observes the basic policy of non-discrimination and affirms the right and freedom of scientists to associate in international scientific activity without regard to such factors as ethnic origin, religion, citizenship, language, political stance, gender, sex or age, in accordance with the Statutes of the International Council for Science.