© 2024
In touch with the world ... at home on the High Plains
Play Live Radio
Next Up:
0:00
0:00
0:00 0:00
Available On Air Stations
KTOT-FM 89.5 serving the northeast TX Panhandle is off the air due to the failure of both air conditioning units needed to cool it's high-power transmitter. The air conditioning units are currently being replaced and other HVAC improvements made. If all goes well, we hope to have these repairs made and KTOT back on the air by the end of the day on Friday, 10/10. We apologize for this this interruption in service. In the meantime, you can always listen on-line through the player above or on HPPR's mobile app to either HPPR Mix, KTOT's regular programming, or HPPR Connect featuring all news and information programming.

Researchers Develop An App To Identify Cattle Through Facial Recognition

ARI SHAPIRO, HOST:

Look at a herd of cows, and their faces might all seem the same even if they are your own cows. So technology is being developed to identify cattle through facial recognition. Seth Bodine of member station KOSU reports that the goal is to track cattle in the event of disease.

SETH BODINE, BYLINE: For Jake Calvert, the only way to tell his cows apart is with colorful ear tags. Take those off, and it gets a little tricky.

JAKE CALVERT: It'd be tough for me to say, oh, well, that's 24, and that's - obviously, the only red cow in the herd, 48 - I'd be able to tell her.

BODINE: But cows have unique faces just like humans. Facial recognition technology can pick up on about 200 key measurement points to identify a human face, what's known as biometrics. It turns out that concept works for cows, too, and artificial intelligence is really good at it.

KC OLSON: The artificial intelligence looked at the pictures for millions of iterations and effectively taught itself which features of the bovine head were most characteristic of the species.

BODINE: That's K.C. Olson. He teaches range livestock nutrition at Kansas State University. He and his team showed the AI lot of cow pictures. Then they played a game. We'll call it, have you seen this cow?

OLSON: Ninety-four percent of the time, artificial intelligence got the right answer.

BODINE: The ultimate goal is to develop an app called CattleTracs. Ranchers would snap a picture of the cow, then send it to a database that Joe Hoagland's company is developing. He says it could lead to a speedier way to track a sick cow.

JOE HOAGLAND: We could trace it and quarantine it and manage it much like we're dealing today with the coronavirus. You know, if you can get on top of it early, you can control it.

BODINE: A highly infectious illness like hoof and mouth disease could all but shut down the cattle industry, disrupting the supply of meat. Right now cows have to be tracked painstakingly through paper documents and sales records to figure out how far a disease has potentially spread. Rosslyn Biggs is a beef extension specialist at Oklahoma State University.

ROSSLYN BIGGS: We need to be able to limit it because undoubtedly, we will have significant interruptions in our supply chain here in the United States. Our export markets will undoubtedly cut us off.

BODINE: The U.S. Department of Agriculture has been eyeing radio electronic ear tags to replace metal tags as official IDs now used widely for cows that travel between states, but these tags are met with resistance from ranchers because of cost. That's why, though Calvert isn't yet sold on the technology, once facial recognition for cows becomes available, he expects ranchers to embrace it.

CALVERT: Because a $9 - even a $20 a year subscription to a phone app is going to be far cheaper than tagging 500 calves with RFID tags.

BODINE: The USDA says it's aware of research into cattle facial recognition and will continue to evaluate it as a possible way to easily identify and trace millions of cows.

For NPR News, I'm Seth Bodine in Oklahoma City.

(SOUNDBITE OF ELKHORN'S "TO SEE DARKNESS") Transcript provided by NPR, Copyright NPR.

Seth Bodine joined KOSU in June 2020, focusing on agriculture and rural issues.