Muzzle biometrics identified harvested red deer with high accuracy

What the study found: The study found that muzzle pattern biometrics can be used to identify harvested wild ungulates, with red deer used as the model species. A comparison of images with the LoFTR (Local Feature TRansformer) method showed individual identification potential.
Why the authors say this matters: The authors say this matters because harvested females are often recorded without verification, and they conclude that biometric methods could provide a verifiable record of harvested game. They also suggest this could support sustainable hunting planning.
What the researchers tested: The researchers took 2,193 photographs from the frontal and overhead directions of 972 harvested red deer during regular game management. They compared the collected images using the LoFTR method and assessed biometric characteristics for automated registration.
What worked and what didn't: The peak identification accuracy was 95.048%. The minimum accuracy was 90.048% when using a combination of overhead and frontal images of high and medium quality. The authors report that the achieved accuracy was around 2% better than comparable recognition systems for pets and livestock.
What to keep in mind: The abstract does not describe detailed limitations beyond the fact that red deer were the model species. It also notes that the comparison was made with pet and livestock recognition systems because no ungulate recognition solution was available.

Key points

• Red deer were used as the model species for testing muzzle pattern biometrics.
• The study analyzed 2,193 photographs from 972 harvested red deer.
• LoFTR-based image comparison showed a peak identification accuracy of 95.048%.
• The lowest reported accuracy was 90.048% with overhead and frontal images of high and medium quality.
• The authors say biometric identification could enable verifiable harvest records and support sustainable hunting planning.