Newsgroups: sci.image.processing
Path: cantaloupe.srv.cs.cmu.edu!rochester!rutgers!uhog.mit.edu!news.mathworks.com!hookup!uwm.edu!cs.utexas.edu!news.sprintlink.net!pipex!bbc!alanr
From: alanr@rd.bbc.co.uk (Alan Roberts)
Subject: Re: help-contrast,brightness
Message-ID: <D5FDqM.KxD@bbc.co.uk>
Sender: usenet@bbc.co.uk
Nntp-Posting-Host: sunf0.rd.bbc.co.uk
Organization: British Broadcasting Corporation, UK
X-Newsreader: TIN [version 1.2 PL0]
References: <3k2dls$1mu@mark.ucdavis.edu>
Date: Tue, 14 Mar 1995 10:13:33 GMT
Lines: 58

Niranjan Avadhanam (avadhana@chervil.ece.ucdavis.edu) wrote:
: hi,

: i have a few simple questions regarding the amplitude non-linearity
: used at the front end of the human visual system models in image
: processing.

: i have noticed that the  models are implemented as a power law
: (b=L^(1/3)) or as a logarithmic law or of the form (b=L/(L+La)).

: Is b brightness? if so why is it mentioned as contrast c in the
: book by Anil Jain (Digital Image Processing)? i am aware of the
: michelson definition of contrast.


: if b is brightness, what are the units? to me they look inconsistent
: because the power law would give a unit based on luminance (cd/m^2)
: whereas the model (L/(L+La)) obviously would give a dimensionless
: unit.

: it would be of great help to me if some knowledgeable netter would
: clarify to me the connection between luminance,brightness,contrast and
: the michelson definition of the contrast.


: should brightness be the input to the following stages of the visual
: system model ie, contrast sensitivity, and the detection process?
: assuming that the contrast sensitivity and detection can be combined
: as a series of filters, would the output of the filterbank give me an
: output in brightness units?

The CIE defines visual brightness as:
 b=116(Y/Yo)^(1/3)-16
where Y is luminance of the object, Yo is luminance of peak white in
the image. The formula is truncated at low luminance levels and replaced
by:
 b=903.3(Y/Yo)

At high brightnesses, the simplification:
 b=L^(1/3)
is not too far off, provided that L is Y/Yo.

Brightness does not have units because it is scaled to the luminance of
the peak white.
 

If you want to know more about this sort of thing, beware, the subject
is very complex, and theories are still evolving. However, you should
look at Charles Poynton's FAQs at:

 elaine.crcg.edu in pub/doc/color

--
************* Alan Roberts **************
* BBC Research & Development Department *
* My views, not necessarily Auntie's    *
*    but they might be, you never know. *
*****************************************
