Computer vision-based augmented visualisation for coffee origins identitation using comprehensive two-dimensional gas...

Area of Science:

• Analytical Chemistry
• Food Science
• Chemometrics

Background:

• Coffee's complex volatile profile is influenced by origin, climate, soil, and processing.
• Hundreds of volatile compounds create analytical challenges for origin identification.
• Existing methods struggle with the high chemical dimensionality of coffee.

Purpose of the Study:

• To develop a robust and rapid method for coffee origin classification.
• To address the analytical challenges posed by coffee's complex chemical matrix.
• To integrate untargeted and targeted analyses for comprehensive characterization.

Main Methods:

• Application of comprehensive two-dimensional gas chromatography (GC×GC) coupled with computer vision (CV).
• Untargeted fingerprinting followed by composite class image generation for origin comparison.
• CV-based peak highlighting, targeted template creation, and multivariate analysis for discriminant compound identification.

Main Results:

• The GC×GC-CV workflow enabled rapid pairwise comparison of coffee origins.
• Differential peaks were identified, leading to the discovery of key discriminant compounds.
• Ion-specific intensity mapping visualized compositional differences across chemical families.

Conclusions:

• The GC×GC-CV workflow provides a powerful platform for coffee chemical characterization and origin classification.
• This integrated approach offers a visually intuitive and robust solution for complex sample analysis.
• The method enhances interpretability and facilitates the identification of origin-specific chemical markers.