Recognizing cattle from their nose prints

Although computer-assisted facial-recognition techniques for cow-indentification are making considerable progress (see previous), there are other ways of doing so – nose prints for example.

Researchers Ary Noviyanto and Aniati Murni Arymurthy of the
Faculty of Computer Science, Universitas Indonesia, Depok, Indonesia, outline the details in a comprehensve report for the journal Computers and Electronics in Agriculture, 99 (2013) 77–84.

“The muzzle pattern from total 48 individuals of two common races of the beef cattle in Indonesia, i.e., Balinese cow and PO cow (hybrid Ongole cow) have been collected. The muzzle patterns have been printed on a piece of paper 30 times for each individual using black ink. Unfortunately, it is very hard to get a good data so that not all of the data can be used. The difficulties of the data capturing include the wet condition of the cattle nose and the cattle’s nervous feeling. The sweat glands of the cattle are quite big, which may leads to the wetness of the nose. The nose wetness leads a smeared muzzle print. In fact, the upper part of the muzzle is wetter compared to the middle part or the lower part. […]

The equipment and materials used in the data capturing are A5-size papers, stamp black ink, cottons and tissues. The procedures of data capturing are presented as follows:
1. The cattle’s head has to be kept still using a rope.
2. Cleaning the nose to eliminate snot using tissues.
3. Once the snot is clear, apply a thin ink layer using cottons on the nose and then print on the paper with upward rolling movement […]”

Their methods, say the authors, show promise when compared to ear-tagging and tattooing. They do point out though, that at the time of writing (2012)

“No method can perfectly recognize each beef cattle correctly.”

See: Beef cattle identification based on muzzle pattern using a matching refinement technique in the SIFT method