Use of animal biometrics for accurate hunting evidence of wild ungulates: red deer as a model species

Overview

Central European wild ungulate populations exhibit continued growth despite hunting pressure, attributable partly to inaccurate harvest documentation and insufficient female removal rates. The study proposes automated biometric identification procedures for recorded game animals, using red deer as a model species to establish verification protocols that ensure accurate harvest registration and improve population management.

Methods and approach

Frontal and overhead photographic images (n=2,193) were collected from 972 harvested red deer during routine game management operations. Individual muzzle pattern biometric analysis was performed using the LoFTR algorithm for automated image comparison and individual identification. Image quality parameters (high, medium) and directional combinations were systematically evaluated to assess recognition accuracy across multiple conditions.

Key Findings

Peak identification accuracy reached 95.048% under optimal conditions. Minimum accuracy of 90.048% was achieved using combined overhead and frontal images of high and medium quality. Performance metrics exceeded comparable recognition systems for domestic livestock and companion animals by approximately 2 percentage points when accounting for dataset size and feature point density.

Implications

Biometric identification via muzzle pattern analysis demonstrates feasibility for automated harvest documentation in ungulate game management. Implementation through mobile application infrastructure capable of transmitting images for computational comparison and identity verification would establish verifiable harvest records and eliminate current documentation gaps arising from unverified harvest claims. Enhanced record accuracy provides quantitative foundation for evidence-based hunting plan adjustments and population control strategies addressing Central European ungulate overabundance.