Related Experiment Video
Updated: Aug 1, 2025
05:56
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
Published on: April 14, 2023
2.6K
A novel method for Pu-erh tea face traceability identification based on improved MobileNetV3 and triplet loss
Zhe Zhang1,2,3, Xinting Yang1,2,3, Na Luo1,3
1National Engineering Research Center for Information Technology in Agriculture, Beijing, 100097, China.
Scientific Reports
|April 28, 2023
Summary
Counterfeiting risks persist despite QR codes and NFC chips. TeaFaceNet, a new model, uses tea face image verification to accurately identify fake Pu-erh tea, enhancing product traceability and market credibility.
Area of Science:
- Agricultural Science
- Computer Science
- Food Science
Background:
- Ensuring Pu-erh tea traceability is vital for quality and safety.
- Current methods like QR codes and NFC chips are vulnerable to counterfeiting.
- Counterfeit products undermine consumer trust and market integrity.
Purpose of the Study:
- To develop an advanced traceability verification model for Pu-erh tea.
- To address the limitations of existing traceability systems against counterfeiting.
- To enhance the credibility and safety of Pu-erh tea products in the market.
Main Methods:
- Proposed TeaFaceNet, a tea face verification model.
- Utilized an improved MobileNetV3 architecture.
- Integrated Triplet Loss for verifying image similarity based on texture features.
Main Results:
- Achieved high recognition accuracies: 97.58% (raw), 98.08% (ripe), and 98.20% (mixed) tea face datasets.
- Demonstrated accurate verification using an optimal threshold.
- Validated the model's robustness and generalization capabilities.
Conclusions:
- TeaFaceNet offers a promising solution for Pu-erh tea traceability.
- The model effectively combats counterfeit products by verifying tea face authenticity.
- Further research can advance Pu-erh tea traceability, quality, and safety assurance.
