Exploring Deep Learning to Predict Coconut Milk Adulteration Using FT-NIR and Micro-NIR Spectroscopy
Agustami Sitorus1, Ravipat Lapcharoensuk1
1Department of Agricultural Engineering, School of Engineering, King Mongkut's Institute of Technology Ladkrabang, Bangkok 10520, Thailand.
Sensors (Basel, Switzerland)
|April 13, 2024
Summary
Deep learning models accurately identified adulteration levels in coconut milk using near-infrared (NIR) spectroscopy. Portable Micro-NIR devices show promise for rapid, on-site food fraud detection.
Area of Science:
- Food Science and Technology
- Analytical Chemistry
- Artificial Intelligence in Spectroscopy
Background:
- Food adulteration poses significant food safety and economic risks.
- Current methods for detecting adulterants like corn flour and tapioca starch in coconut milk are often slow and lack robustness.
- Developing rapid, accurate, and non-destructive analytical techniques is crucial for effective adulteration detection.
Purpose of the Study:
- To explore the efficacy of deep learning algorithms for quantifying coconut milk adulteration.
- To compare the performance of benchtop FT-NIR and portable Micro-NIR spectrophotometers for adulterant detection.
- To develop a robust prediction model for identifying adulteration levels using near-infrared (NIR) spectral data.
Main Methods:
- Coconut milk samples were deliberately adulterated with corn flour and tapioca starch (1-50%).
- Near-infrared (NIR) spectra were acquired using both FT-NIR and portable Micro-NIR spectrophotometers.
- Four modified deep learning architectures (CNN, S-AlexNET, ResNET, GoogleNET) were applied to the NIR datasets for quantitative analysis.
Main Results:
- Deep learning models demonstrated reliable performance in predicting adulteration levels (R²: 0.886-0.999, RMSE: 0.370-6.108%).
- Ratio of Percent Deviation (RPD) values indicated excellent quantitative prediction capabilities across most models and instruments.
- The portable Micro-NIR showed greater promise than FT-NIR for detecting solid adulterants, suggesting potential for in situ applications.
Conclusions:
- Deep learning algorithms combined with NIR spectral data offer a rapid, accurate, and non-destructive method for evaluating coconut milk adulterants.
- The study validates the feasibility of using advanced computational techniques for food quality control and fraud prevention.
- Portable NIR technology, particularly Micro-NIR, is a promising tool for on-site, real-time food adulteration monitoring.
More Related Videos
Related Concept Videos
01:03IR Frequency Region: Fingerprint Region
875
IR spectra are divided into two main regions: the diagnostic region and the fingerprint region. The diagnostic region of the spectrum lies above 1500 cm−1. The absorptions resulting from single-bond vibrations of the N–H, C–H, and O–H stretch at higher wavenumbers and appear on the left side of the spectrum. The stretching absorptions of the C≡C and C≡N occur between 2100–2300 cm−1. In contrast, those arising from stretching absorptions of the...
875
01:29IR and UV–Vis Spectroscopy of Aldehydes and Ketones
5.5K
Infrared spectroscopy, also known as vibrational spectroscopy, is mainly used to determine the types of bonds and functional groups in molecules. In aldehydes and ketones, the carbonyl (C=O) bond shows an absorption around 1710 cm-1. The C=O bond vibration of an aldehyde occurs at lower frequencies than that of a ketone. In addition to the C=O absorption in an aldehyde, the aldehydic C–H bond also gives two peaks in the 2700–2800 cm-1 range. This absorption, coupled with the...
5.5K
01:28IR and UV–Vis Spectroscopy of Carboxylic Acids
4.0K
In IR spectroscopy of carboxylic acids, the C=O bond shows a characteristic band between 1710 and 1760 cm⁻¹, and the O–H bond exhibits a broad band between 2500 and 3300 cm⁻¹.
However, the stretching absorptions for the C=O bond vary depending on the structure of carboxylic acids. The C=O bond of the free carboxylic acids shows a higher stretching frequency, 1760 cm−1, while H-bonded carboxylic acids (dimers) exhibit stretching absorptions at a lower frequency,...
However, the stretching absorptions for the C=O bond vary depending on the structure of carboxylic acids. The C=O bond of the free carboxylic acids shows a higher stretching frequency, 1760 cm−1, while H-bonded carboxylic acids (dimers) exhibit stretching absorptions at a lower frequency,...
4.0K
01:11Applications of IR Spectroscopy: Overview
587
The non-destructive nature and ability to provide valuable chemical information make IR spectroscopy a versatile technique with broad applications in various scientific and industrial fields. IR spectroscopy is commonly used to identify and characterize organic and inorganic compounds. It provides information about the functional groups present in a molecule and the bonding between atoms. This helps in the structural elucidation of compounds during organic synthesis, pharmaceutical research,...
587
01:26Spectroscopy of Carboxylic Acid Derivatives
2.3K
Infrared spectroscopy is primarily used to determine the types of bonds and functional groups. In carboxylic acid derivatives, a typical carbonyl bond absorption is observed around 1650–1850 cm−1. For esters, the absorption is recorded at around 1740 cm−1, while acid halides show the absorption at about 1800 cm−1. Another acid derivative, the acid anhydrides, exhibit two carbonyl absorption around 1760 cm−1 and 1820 cm−1, arising from the symmetrical and...
2.3K
01:15NMR Spectroscopy and Mass Spectrometry of Aldehydes and Ketones
4.0K
In aldehydes, the hydrogen atom connected to the carbonyl carbon helps distinguish aldehydes from other carbonyl compounds using ¹H NMR spectroscopy. The closeness of aldehydic hydrogen to the electrophilic carbonyl carbon highly deshields the hydrogen atom causing its signal to appear around 10 ppm in the ¹H NMR spectra. α hydrogens split the aldehydic proton signal, which helps identify the number of α hydrogens in the molecule. For instance, one α hydrogen creates a...
4.0K