Difference between revisions of "Automated photo identification"

From Beam Reach Wiki
Jump to: navigation, search
(qDdpFTwPkNrbfuW)
 
(2 intermediate revisions by 2 users not shown)
Line 1: Line 1:
Super ifnoarmtive writing; keep it up.
+
Beam Reach has an [http://beamreach.org/gallery extensive archive of photographs].  Many of these could be used retrospectively to identify individuals and assess their condition.  Some potentially useful photographs have be tagged with "id."  Programs that may be useful for processing these images are listed below.
 +
 
 +
* [http://www.europhlukes.net/ Europhlukes], Phlex and Match (used to match sperm whales between Azores and Norway, Marmam, Oct. 2008; more general discussion of [http://homepages.cwi.nl/~ely/projects.htm europhlukes applications])
 +
* Darwin (automatic characterization and matching to database of bottlenose dolphins) [http://darwin.eckerd.edu/ home page] | [http://beamreach.org/wiki/images/9/9d/ScottHale_src_gf06.pdf PDF of automatic algorithm]
 +
* FinSean
 +
 
 +
 
 +
* A reference to get:
 +
<nowiki>Shape descriptors computed from photographs of dolphin dorsal finsfor use as database indices
 +
Hillman, G.R.; Tagare, H.; Elder, K.; Drobyshevski, A.; Weller, D.; Wursig, B.
 +
Engineering in Medicine and Biology Society, 1998. Proceedings of the 20th Annual International Conference of the IEEE
 +
Volume 2, Issue , 29 Oct-1 Nov 1998 Page(s):970 - 973 vol.2
 +
Digital Object Identifier  10.1109/IEMBS.1998.745609
 +
Summary:Field photographs of bottlenosed dolphins (Tursiops truncatus) are used for identification of individuals based on notches that are present in the fin; comparison by eye with a collection of photographs is commonly used but is cumbersome. A system is reported here that extracts from photographs a set of features describing the pattern of notches in the fin; these features will be used to index an image database. A method based on modeling the fin edge with an affine invariant Fourier transform was tested but was not successful. A method based on notch location with a stiff snake followed by Gaussian curve fitting, however, produced a promising set of features</nowiki>

Latest revision as of 12:12, 11 May 2012

Beam Reach has an extensive archive of photographs. Many of these could be used retrospectively to identify individuals and assess their condition. Some potentially useful photographs have be tagged with "id." Programs that may be useful for processing these images are listed below.


  • A reference to get:

Shape descriptors computed from photographs of dolphin dorsal finsfor use as database indices Hillman, G.R.; Tagare, H.; Elder, K.; Drobyshevski, A.; Weller, D.; Wursig, B. Engineering in Medicine and Biology Society, 1998. Proceedings of the 20th Annual International Conference of the IEEE Volume 2, Issue , 29 Oct-1 Nov 1998 Page(s):970 - 973 vol.2 Digital Object Identifier 10.1109/IEMBS.1998.745609 Summary:Field photographs of bottlenosed dolphins (Tursiops truncatus) are used for identification of individuals based on notches that are present in the fin; comparison by eye with a collection of photographs is commonly used but is cumbersome. A system is reported here that extracts from photographs a set of features describing the pattern of notches in the fin; these features will be used to index an image database. A method based on modeling the fin edge with an affine invariant Fourier transform was tested but was not successful. A method based on notch location with a stiff snake followed by Gaussian curve fitting, however, produced a promising set of features