The National Science Foundation has made a $US256,000 grant to Wildlife Protection Management (WPM) it will use to work on a facial recognition system to aid in population control in wild horses.
Previously, WPM had introduced a prototype system that remotely implants radio-frequency identification, or RFID, chips, vaccines and contraceptives into horses.
WPM CEO Roch Hart said the RFID system, which is attached to an alfalfa feeding station, had been a success: “Better than we had hoped.” He said using facial recognition technology would cut costs by not having to implant every horse with a chip.
“We do hope to take this technology for horses to other species and scale that technology up. This grant certainly helps that traction,” Hart said.
New Mexico State University’s entrepreneur and business incubator Arrowhead Center helped WPM define the need for facial recognition technology. The grant will help validate the feasibility of using facial recognition alongside the RFID for the identification of not only wild horses but also feral pigs and deer.
More assistance for WPM will come through the collaboration with New Mexico State University’s College of Engineering Associate Professor Laura Boucheron. With the support of a graduate student, Boucheron will spend a year dissecting WPM’s data and creating an effective method of animal recognition through videos and images.
“While they have RFID chips, the horses might have recognizable patterns on their faces or flanks or even scars that we can use to build an algorithm that detects individual animals,” Boucheron said.
“The ultimate goal is to recognize individual animals based on videos or images and match them up with their RFID identification.
“What we do can have a direct impact on a problem with the horse population. It’s something that has quite the real-world applicability,” Boucheron said.
Throughout the United States, wild horses still roam through the wide-open spaces. While they may be wild and free, they come at a cost to taxpayers, as much as $1 billion from 2019 to 2023, according to the US Bureau of Land Management.