[Taxacom] Automated species Identification by Image Analysis

[Les Watson]

The approach has long had specialised applications (e.g., in medical 
diagnosis), but I doubt whether it has much general taxonomic potential.

I conducted trials in the early 1970s with the originators of an early 
Image Analyzer (Imanco, of Cambridge), with a view to identifying grass 
epidermal preparations. These should be espcially amenable to this 
approach, because preparations are easily orientated consistently for 
comparative analyses, and much can be achieved via 2-dimensional views 
of slides and even from photos.  It proved a fizzer, because (a), one 
had to laboriously provide precise definitions of the "obvious" 
differences between straight and tesselated cell walls, between nodular 
and crenate silica-bodies, etc; and (b), automation would involve 
(laboriously) teaching the system not only how to make informed 
approximations, but how to recognise and ignore rubbish (distorted 
material, fractures silica-bodies, dirt, etc.).

Image analytical techniques have improved, but the future of automated 
identification surely lies in comparative biochemical analyses of 
proteins, and ultimately of DNA. Precise identification of airborne 
grass pollens (important re allergies, and impossible microscopically) 
would be a good place to pursue this - if indeed it isn't already being 
done .......

-----------------------------------------------------------
Les Watson
10 Maitland Avenue, Little Grove, Albany WA 6330, Australia
Email: leswatson at westnet.com.au  Phone: +61 (8) 98 44 4398

---------------------------------------------------------=-

David Remsen (GBIF) wrote:
> Rich,
> 
> Recall my brothers work on image analysis for gross grouping of  
> zooplankton into major groups.  It's not species but I suspect for  
> many groups you won't get species-level diagnosis but there is  
> probably a lot of potential in the higher sorting by images that he  
> was doing.   I think they were getting pretty good results.
> 
> David
> 
> 
> On Sep 7, 2009, at 3:55 AM, Richard Pyle wrote:
> 
> 
>>I have been asked by a non-biologist colleague for information on  
>>the use of
>>automated algorithms for species identification from digital  
>>images.  I
>>recall several such projects (one of them was called "DAISY"), and I  
>>believe
>>there was some success with vein patterns in bee wings, and perhaps  
>>other
>>examples as well.  Does anyone have any insights on this, and  
>>whether it's
>>seen a a feasible technology worthy of further investment of time and
>>effort? How much effort/investment has already gone into it?
>>
>>Thanks in advance...
>>
>>Aloha,
>>Rich
>>
>>Richard L. Pyle, PhD
>>Database Coordinator for Natural Sciences
>> and Associate Zoologist in Ichthyology
>>Department of Natural Sciences, Bishop Museum
>>1525 Bernice St., Honolulu, HI 96817
>>Ph: (808)848-4115, Fax: (808)847-8252
>>email: deepreef at bishopmuseum.org
>>http://hbs.bishopmuseum.org/staff/pylerichard.html
>>
>>
>>
>>
>>_______________________________________________
>>
>>Taxacom Mailing List
>>Taxacom at mailman.nhm.ku.edu
>>http://mailman.nhm.ku.edu/mailman/listinfo/taxacom
>>
>>The Taxacom archive going back to 1992 may be searched with either  
>>of these methods:
>>
>>(1) http://taxacom.markmail.org
>>
>>Or (2) a Google search specified as:  site:mailman.nhm.ku.edu/ 
>>pipermail/taxacom  your search terms here
>>
> 
>