Muzzle Net

Deep Learning Based Cattle Identification using Muzzle Pattern Recognition

MuzzleNet is a deep learning project focused on biometric identification of cattle using unique muzzle patterns. Similar to human fingerprints, every cow has a distinct muzzle texture, making it a reliable biometric feature for livestock identification.

---

Current Results

Example similarity scores from testing:

Same Cow

Mean: 0.935258
Min : 0.24305052
Max : 0.99367917

Different Cow

Mean: 0.1563636
Min : -0.57475066
Max : 0.9371778

Similarity Distribution

---

Research Reference

Reference Paper

Deep Learning-Based Cattle Identification Using Muzzle Pattern Images

Cattle Identification using Muzzle Images and Deep Learning Techniques

---

Dataset Used

The project uses Zenodo Beef Cattle Muzzle Dataset

Download Dataset