Non-destructive freshness assessment of mackerel (Scomber japonicus) using colorimetric analysis and machine...

Area of Science:

• Food Science
• Analytical Chemistry
• Machine Learning

Background:

• Seafood quality assessment is crucial but traditional methods are slow and destructive.
• Objective, non-destructive methods are needed for real-time quality control.
• Color changes in fish are indicators of degradation.

Purpose of the Study:

• To evaluate colorimetric analysis as a non-destructive method for assessing mackerel freshness.
• To correlate color changes with microbiological and physicochemical quality indicators.
• To develop predictive models for seafood freshness using color data and machine learning.

Main Methods:

• Colorimetric analysis (RGB, HSV, Lab) of mackerel eye, belly, and dorsal regions.
• Storage of mackerel at 4 °C and 10 °C.
• Microbiological (viable cell count, total coliforms) and physicochemical (pH, TVBN) assessments.
• Development of predictive models using Multivariate Linear Regression (MLR) and Support Vector Regression (SVR).

Main Results:

• Progressive darkening of mackerel color observed during storage.
• Color changes correlated with increased microbial load, pH, and total volatile basic nitrogen (TVBN).
• SVR models showed superior prediction accuracy for non-linear freshness indicators like pH and total coliforms.

Conclusions:

• Colorimetric analysis combined with machine learning provides a viable non-destructive method for evaluating seafood freshness.
• This approach can be integrated into industrial quality control systems for real-time assessment.
• The study highlights the potential for objective and efficient seafood quality monitoring.