TMU Develops AI Egg Freshness Scanner

By combining hyperspectral imaging with deep learning-based wavelength selection, the research team successfully created a new approach to egg quality assessment. The technology is the result of collaboration between a research team led by Associate Professor Yung-Kun Chuang from the TMU School of Food Safety and Professor Pauline Ong from the Faculty of Mechanical and Manufacturing Engineering, Universiti Tun Hussein Onn Malaysia (UTHM).

The research findings were published in Current Research in Food Science, an international journal with a 2024 impact factor of 7.0, ranking in the top 9.9% of the Food Science & Technology category. The publication highlights TMU’s growing influence in global food safety research and advanced AI applications.

Associate Professor Yung-Kun Chuang, School of Food Safety, College of Nutrition, Taipei Medical University

Eggs are one of the most vital and complete protein sources in our daily diet—rich in nutrients, vitamins, and minerals. However, the freshness of eggs, which is key to their nutritional value, safety, and overall quality, can be compromised during storage, transportation, and distribution. Traditional evaluation methods rely on destructive laboratory testing, making them unsuitable for real-time or large-scale monitoring along the production and supply chain. This has driven the need for rapid, accurate, and non-destructive technologies that can assess egg quality without breaking the shell.

Research Workflow and Data Analysis Summary

To overcome these limitations, the research team developed a novel wavelength selection technique that integrates distance correlation analysis with a convolutional neural network model. This approach analyzes hyperspectral images of eggs to reduce spectral variability, identify the most informative wavelengths, and to enhance the accuracy of freshness levels. Compared with commonly used wavelength selection regression models, this innovative method demonstrated stability, reliability and overall performance, offering strong potential for real-world industrial adoption.

The achievements of this international collaboration pointed a new era in egg quality control. The AI-powered hyperspectral egg scanner provides fast and non-destructive testing, and provides quantifiable freshness assessments, and adaptable across production, processing, and distribution.

assessments, and adaptable across production, processing, and distribution. 111modernization of the egg industryalso reinforces TMU’s leadership in food safety research and artificial intelligence applications.intelligence applications.