Classification of organic beef freshness using VNIR hyperspectral imaging.

Area of Science:

• Food Science
• Analytical Chemistry
• Spectroscopy

Background:

• Consumer trust in the food industry hinges on accurate meat product labeling.
• Verifying meat authenticity is crucial for producers, retailers, and consumers.
• Hyperspectral imaging and CIELAB measurements offer potential for objective meat classification.

Purpose of the Study:

• To develop and evaluate methods for classifying beef based on its condition (fresh, frozen-thawed, matured).
• To investigate the utility of hyperspectral imaging in the 500-1010nm waveband for beef classification.
• To assess the effectiveness of CIELAB color measurements in differentiating beef states.

Main Methods:

• Hyperspectral imaging was employed to capture spectral data from beef samples.
• CIELAB color measurements were taken to quantify color attributes.
• Classification models were developed using spectral and colorimetric data to distinguish between fresh, frozen-thawed, and matured beef.
• The performance of reduced spectral models was also evaluated.

Main Results:

• High Correct Classification Rates (CCR) were achieved using CIELAB for fresh vs. frozen-thawed (0.93) and fresh vs. matured (0.92) beef.
• Perfect classification (CCR=1.00) was obtained for matured vs. matured frozen-thawed beef using the full spectral range.
• CIELAB was effective for most comparisons, except between matured and matured frozen-thawed beef.

Conclusions:

• CIELAB color coordinates are valuable for classifying beef states, particularly for differentiating fresh from altered conditions.
• Hyperspectral imaging, especially across the 500-1010nm range, provides robust data for distinguishing subtle changes in beef.
• The study highlights the potential of these non-destructive techniques for ensuring meat authenticity and quality assurance.