Issue 255: Shape

Starting Date: 
Working Group: 
Closing Date: 

Posted by Thanasis Velios on 31/7/2014 

Hello list, 

I am looking for some advice regarding describing the shape of objects. How is shape considered in the CRM? For example this object: 
has an "oval hole in base". The hole is an E26 Physical Feature. Would you simply use E55 Type as in: 


oval holes 
round holes 
or is there a more elegant way of doing it? 

Posted by Vladimir on 5/8/2014 
Hi Thanasis!

How is shape considered in the CRM?

There's no special modeling of shapes.

E55 Type as in: 
   <holes by shape> 
     oval holes 
     round holes 

or is there a more elegant way of doing it? 

If you have only a few of those then go ahead with it. 

If there are many combinations, better split them up, e.g.: 
   P2X_has_area_shape subPropertyOf P2_has_type. 
   P2Y_has_volume_shape subPropertyOf P2_has_type. 
and then 

   <feature1> P2X_has_area_shape <thes/shape/area/oval>; 
     P2Y_has_volume_shape <thes/shape/volume/hole>. 
   <feature2> P2X_has_area_shape <thes/shape/area/square>; 
     P2Y_has_volume_shape <thes/shape/volume/bump>.

Current Proposal: 

In 31st joined meeting of the CIDOC CRM SIG, ISO/TC46/SC4/WG9 and the 24th FRBR - CIDOC CRM, the crm-sig reviewing this issue decided that FAQ on shape and color required

Heraklion, Crete, October 2014


In the 32nd joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 25th FRBR - CIDOC CRM  discussed that the shapes are morphological classes. The problem is the colours you may observe are depended by the granularity you look up.  Discussing about the physical parameters that  the colour depends on, we decided that these are the  normalized light source and  the surface area selected integrade the light reflections. (HIS dimensions relates to RTI imaging).

The crm-sig assigned to MD to write and Steve Stead to review the FAQ on shape and color

In 34th Joined meeting the crm-sig accepted the text about the FAQ "what is Color? The text is

"Colour is a property which can be defined in the sense of physics or in the sense of human neurophysiology. The way humans perceive colors is complicated, but scientifically well understood.

Color of a surface in the sense of physics is the spectrum of light frequencies emitted from a surface. Neglecting active light sources, it is a reaction on the frequencies of the incoming light, and therefore depends on it. Out of physiological reasons, standard colors are defined with respect to a standard light source that represents an ideal daylight. Such devices are used in labs. The standard daylight source has a continuous spectrum of a known shape.

The human eye registers, i.e., projects out of the reflected spectrum three sub-spectra, one centering at red, one at green, and one at blue. The combination of the intensities of these three components produces our sensory impressions of all colors, including violet and yellow. These effects can be reproduced by adequate mechanical sensors and processing.

The surface itself may have unlimited details, microscopic differences in reflection. The reflected light is always the sum of the light reflected from some surface of a certain size. Therefore color is only defined with respect to a certain spot size. Seeing a surface from a certain distance increases the spot size, and hence may produce new integral colors, as well known from a TV screen surface. Also quick change of light ( >25 cps) sums up to an integral sensory reaction.

In addition, our brain has two extraordinary processing functions, which keep our color impression much more constant than physics would make us expect:

A)     White balance: Standard daylight is “white”. Our brain detects deviations of the light source from being white, and corrects the sensory impression very widely in order to produce correct colors as if the source were white. Digital cameras try more or less successfully to simulate this process as “automatic white balance”.

B)     Surface homogeneity: Our brain mysteriously detects shadows and colored reflections on homogeneously colored surfaces and presents itself the corrected surface color. So we recognize for instance a pullover to be uniformly red within complex light conditions, reducing the surface color to the woolen thread color.

Consequently, describing “the colour” of an object is not a question of a simple vocabulary. Even given under standard light conditions, it still needs the definition of the spot size of integration, and the representative surface parts of features meant by this categorization. 

Art conservators recently tend to scan art objects by much more colors than the three colored subspectra the human eye distinguishes. This allows for much better print representation and scientific analysis of colorants. "

Reference to Issues:

Meetings discussed: