Related Experiment Video
Updated: Jun 20, 2025
09:15
DNA-based Fish Species Identification Protocol
Published on: April 28, 2010
29.4K
What the fish? Tracing the geographical origin of fish using NIR spectroscopy
Nidhi Dalal1, Raffaela Ofano1, Luigi Ruggiero1
1Department of Agricultural Sciences, University of Naples 'Federico II', Italy.
Current Research in Food Science
|July 18, 2024
Summary
Near-Infrared (NIR) spectroscopy is a powerful tool for authenticating fish and combating food fraud. This review covers sampling, pre-processing, and advanced machine learning analysis for reliable fish identification.
Area of Science:
- Food Science
- Analytical Chemistry
- Spectroscopy
Background:
- Food authentication is crucial due to complex supply chains and the prevalence of fish fraud.
- Near-Infrared (NIR) spectroscopy offers an efficient method for verifying fish authenticity.
Purpose of the Study:
- To review recent advancements in NIR spectroscopy for fish authentication.
- To cover the entire process from sample preparation to data analysis.
Main Methods:
- Detailed examination of sampling strategies for representative traceability studies.
- Emphasis on NIR spectra pre-processing techniques to handle overlapping bands and scattering.
- Review of unsupervised (e.g., PCA) and supervised (e.g., LDA, PLS-DA) multivariate analysis methods.
- Exploration of machine learning approaches for fish authentication modeling.
Main Results:
- Effective pre-processing is critical for successful NIR spectral analysis and model development.
- Unsupervised analysis can aid in feature reduction for complex NIR data.
- Machine learning models show significant promise, potentially bypassing the need for preliminary unsupervised analysis.
Conclusions:
- NIR spectroscopy, coupled with appropriate data analysis, is a robust technology for fish authentication.
- Advancements in pre-processing and machine learning enhance the reliability and efficiency of NIR-based fish fraud detection.
