StripeSpotter

StripeSpotter is an automatic individual animal identification system for animals with prominent stripes or patches, developed as a joint project between the http://compbio.cs.uic.edu/'>Computational Population Biology laboratory at the University of Illinois at Chicago, and the http://www.princeton.edu/~dir'>Equid Research and Conservation laboratory at Princeton University. It is intended to be used to identify animals in the wild, and to build biometric databases using photographs taken in the field. We are currently using it to build a zebra-print database for Plains and Grevys zebra in Kenya.

It has a number of friendly features:

• Cutting edge algorithms
  
  • CO-1 and our own StripeCode algorithm (see below)
    
  • Written in C++ for speed
    
  • Extremely efficient: runs fast even on old laptops

• Database is stored in plain-text, Excel-compatible CSV format
  
  • Easy backups (e.g., just drag the whole data folder to a USB key)
    
  • Images are stored on disk and can be browsed like any other folder
    
  • Can use Dropbox or similar service to synchronize database between many users
    (credit for idea: Siva Sundaresan)

• Allows the creation of databases with ecological metadata
  
  • e.g., GPS coordinates, animal demographics, sighting notes
    
  • Add extra fields directly to the CSV spreadsheet database

• Completely free and open-source
  • Full transparency, and the price is right!
    
  • Runs on Windows, Mac and Linux
    
  • No messy installers - just unzip the download and run!

For more details on the StripeCode algorithm, please refer to the http://compbio.cs.uic.edu/~mayank/papers/LahiriEtal_ZebraID11.pdf'>ACM ICMR 2011 conference paper. If you use StripeCode or our dataset in academic research, please cite the following paper:

• M. Lahiri, C. Tantipathananandh, R. Warungu, D.I. Rubenstein, T.Y. Berger-Wolf. Biometric Animal Databases from Field Photographs: Identification of Individual Zebra in the Wild. Proceedings of the ACM International Conference on Multimedia Retrieval (ICMR 2011), Trento, Italy, 2011.

Bibtex:

@InProceedings{lahiri11_biometric,


  title={Biometric Animal Databases from Field Photographs: Identification of Individual Zebra in the Wild},


  author={M. Lahiri and C. Tantipathananandh and R. Warungu and D.I. Rubenstein and T.Y. Berger-Wolf},


  year={2011},


  booktitle={Proceedings of the ACM International Conference on Multimedia Retrieval (ICMR 2011)},


  publisher={ACM Press}


}

Contact:

mlahiri at gmail dot com

Jump to a wiki page:

• Download and install
• Usage walkthrough
• Sample zebra dataset
• Reproducing results from the ACM ICMR 2011 paper

Acknowledgements

This work is part of a project performed in the joined Princeton-UIC
Computational Population Biology Course in Spring 2010
(http://compbio.cs.uic.edu/~tanya/teaching/KenyaCourse.html'>http://compbio.cs.uic.edu/~tanya/teaching/KenyaCourse.html), with
co-instructors Tanya Berger-Wolf (University of Illinois at Chicago),
Daniel Rubenstein and Iain Couzin (Princeton University), who were
instrumental in several parts of this research. We thank he Kenya
Ministry of Education, Science and Technology (research permit MOST
13/001/29C 80Vol.11 to D.I. Rubenstein), the staff at
Mpala Resarch Centre, Kenya and fellow graduate students at
EEB-Princeton University and CS at University of Illinois at Chicago.
Funding was provided by Department of Ecology and Evolutionary Biology
of Princeton University, generous contribution by Bill Unger (for the
UIC students in the course), UIC College of Engineering, Department of
Computer Science at UIC, UIC Graduate Research Award (Lahiri), NSF
III-CXT 0705311 (Rubenstein) and
IIS-CTX-0705822 and NSF IIS-CAREER-0747369 (Berger-Wolf).

See press coverage about StripeSpotter:

http://www.msnbc.msn.com/id/42500680/ns/technology_and_science-science/'>http://www.surftilyoudrop.com/images/logoMSNBC.jpg' /> http://www.newscientist.com/blogs/onepercent/2011/04/barcode-scanner-for-zebras.html'>http://www.newscientist.com/img/misc/ns_logo.jpg' /> http://www.popsci.com/technology/article/2011-04/barcode-scanner-zebras-helps-biologists-track-individual-animals'>http://www.popsci.com/files/icon-follow-magplus-80x80.jpg' /> http://www.wired.co.uk/news/archive/2011-04/8/zebra-barcode-scanner'>http://www.wefeelfine.org/common/news/wired_logo.gif' /> http://www.readwriteweb.com/archives/at_long_last_zebras_are_scanable.php'>http://www.logodesignworks.com/blog/images/read-write-web-logo.jpg' /> http://news.discovery.com/tech/barcode-scanner-identifies-striped-animals-110407.html'>http://news.discovery.com/images/tophat/discovery-news-logo.png' />

And some of our favorite mentions:

NPR Wait Wait Don't Tell me (Starting from minute 2:48)
http://www.npr.org/player/v2/mediaPlayer.html?action=1&t=1&islist=false&id=135453443&m=135453421'>http://www.npr.org/player/v2/mediaPlayer.html?action=1&t=1&islist=false&id=135453443&m=135453421

Transcript:
http://www.npr.org/2011/04/16/135453443/limericks'>http://www.npr.org/2011/04/16/135453443/limericks

Telling zebras apart is a hard road


and I'm making mistakes by the car load.


So I got this device that keeps track of their stripes


and just scans them like they are a ... BARCODE!

NPR's Marketplace TechReport (tacked on at the end of the piece on
online privacy, starting at minute 3:17):
http://marketplace.publicradio.org/display/web/2011/04/13/tech-report-what-are-kerry-mccain-privacy/'>http://marketplace.publicradio.org/display/web/2011/04/13/tech-report-what-are-kerry-mccain-privacy/