A Data Science Approach to Colorimetric and pH Analysis for Freshness Monitoring

Polysaccharide-based films represent a sustainable alternative for the development of intelligent packaging systems capable of monitoring food quality through colorimetric changes. In this study, films formulated with chitosan, starch, glycerol, and Hibiscus sabdariffa extract were evaluated as potential freshness indicators for pork meat. The films were prepared using different extract concentrations and stored at 6 and 25 °C before and after contact with pork samples for 36 h. Their colorimetric response was quantified through digital image analysis using the RGB color model combined with the K-means clustering algorithm. The relationships among the experimental variables were investigated using descriptive statistics, Spearman correlation analysis, and feature relevance metrics, including Information Gain, Gain Ratio, and Analysis of Variance (ANOVA). The results demonstrated that polymer blend concentration was the most influential factor in differentiating film performance, whereas storage time and extract concentration significantly affected both the colorimetric response and pH values. Multivariable analysis revealed distinct patterns associated with film stability and the physicochemical changes induced by contact with pork meat during storage. These findings highlight the potential of integrating polysaccharide-based materials with data science approaches to support the design, optimization, and performance evaluation of intelligent packaging systems for real-time freshness monitoring of meat products. Beyond the specific application evaluated, the proposed methodology demonstrates how data science tools can support the objective assessment of intelligent biodegradable films, facilitating their future implementation in real-time food quality monitoring systems.