Milk adulteration identification using hyperspectral imaging and machine learning
Muhammad Aqeel1, Ahmed Sohaib1, Muhammad Iqbal2
1Advanced Image Processing Research Lab (AIPRL), Institute of Computer and Software Engineering, Khwaja Fareed University of Engineering and Information Technology, Rahim Yar Khan 64200, Pakistan.
Journal of Dairy Science
|November 9, 2024
Summary
This study developed a 100% accurate method to detect and categorize milk adulteration using hyperspectral imaging and machine learning. The findings offer a practical solution for ensuring milk quality and consumer safety globally.
Area of Science:
- Food Science
- Analytical Chemistry
- Machine Learning
Background:
- Milk adulteration is a significant global issue, particularly in regions with weak monitoring systems.
- Adulterated milk poses severe health risks, including potentially fatal diseases.
- Accurate detection and categorization of milk adulteration are vital for consumer safety and the dairy industry.
Purpose of the Study:
- To develop and validate methods for detecting and categorizing milk adulteration.
- To compare destructive and nondestructive analytical techniques for milk quality assessment.
- To establish a highly accurate, user-friendly system for identifying milk adulterants.
Main Methods:
- Destructive analysis using the Lactoscan system for parameters like fat, protein, and lactose.
- Nondestructive analysis employing hyperspectral imaging (HSI) for spectral signature extraction.
- Machine learning algorithms, including Linear Discriminant Analysis (LDA), trained on a milk adulteration dataset.
Main Results:
- Linear Discriminant Analysis (LDA) demonstrated superior performance in identifying milk adulteration.
- The proposed pipeline achieved 100% validation accuracy in detecting and categorizing milk adulterants.
- The study successfully established a multiclass model for milk adulterant behavior detection.
Conclusions:
- Hyperspectral imaging combined with machine learning provides an effective, non-destructive method for milk adulteration detection.
- The developed model offers significant practical applications for real-time milk quality assessment.
- This research contributes a robust solution to address the global challenge of milk adulteration.
