An Advanced Colour Calibration Method for Fish Freshness Assessment: a Comparison Between Standard and Passive Refrigeration Modalities

摘要

Freshness represents a pivotal aspect in fish product for both security and quality. Its evaluation still represents the key factor driving the consumer' choices. Fish appearance is affected by many different factors that demand the contribution of different disciplines to be understood: from the physical and optical properties to the slaughtering and post-slaughtering conditions. An innovative preservation system is represented by the Passive Refrigeration PRS (TM) developed for the preservation and transport of perishable food products. Scientific methods for product freshness evaluation may be conveniently divided into two categories: sensorial and instrumental. In this study, an instrumental method of colour calibration and discrimination is proposed at pilot scale for automatic evaluation of gilthead seabream (Sparus aurata) freshness. We propose a non-destructive method based on the colorimetric imaging of the whole external body of seabreams to evaluate through multivariate partial least squares which approach the differences in the freshness preservation under four refrigeration modalities. The matrix of the independent variables is represented by RGB values for each pixel belonging to an extracted region of interest (129,633 values). The dependent variable is composed by two dummy variable corresponding to fresh (T-0) or non-fresh (T-2) individuals. T-1 individuals were used as external test. The results quantified significant colorimetric differences between fresh and non-fresh fish. All fish used to create the model (T-0 and T-2) were correctly classified as fresh or non-fresh, while external test individuals (T-1) were all classified as fresh. The proposed imaging method merges different image analysis techniques: (a) colorimetric calibration, (b) morphometric superimposition and (c) partial least square discriminant analysis modelling. This innovative and non-destructive approach allows the automatic assessment of fish freshness.