Skip to content

TigerNet Identified Individual Amur Tigers From Side-View Images With 95–96.5% Accuracy

Amur tiger showing its side stripe pattern at Siberian Tiger Park in Harbin, China

TigerNet, a computer-vision system designed to recognize individual Amur tigers (Panthera tigris altaica) from their stripe patterns, reached 95% accuracy on left-side photographs and 96.5% on right-side photographs. The study, published on 7 January 2026, tested the system on 12,625 tiger images from Siberian Tiger Park in Harbin, China.

The method was inspired by fingerprint recognition. First, it enhances stripe geometry and orientation so that the pattern stands out more clearly from the background. A modified EfficientNetV2Small then extracts distinguishing visual features, and a similarity-verification stage compares stripe angles and orientations while accounting for changes in pose.

TigerNet outperformed the comparison methods used in the study by 2.5–5.7 percentage points on the tiger data. The authors also found that enhancing the stripe pattern reduced recognition time by 37%.

To test whether the method was tied specifically to tiger stripes, the researchers also applied TigerNet to a public dataset of wild zebras. It reached 90.2% accuracy and exceeded MiewID by 5.7 percentage points. The authors therefore present TigerNet as a non-invasive image-based identification method with potential use for other species that have individually distinctive body patterns.

About this content: This story was produced with AI assistance within an editorial workflow developed by Wildlife Vagabond. Editorial responsibility remains with Wildlife Vagabond.How AI is used

Other research

SPECIES IN THIS STORY

Species in this story

Tiger Panthera tigris Explore species

Independent research and conservation news archive

Did you find this information useful?

Wildlife Vagabond is independently built and maintained. Voluntary support helps cover source verification, hosting and continued work on research and conservation news.

Support Wildlife Vagabond

The news archive will remain freely available.