Academician Sun Dawen of South China University of Technology published a review in the top international food journals, explaining the application progress of hyperspectral imaging in fast and non-destructive sensory analysis of food

Recently, Academician Sun Dawen of South China University of Technology published an online publication titled "Rapid and noninvasive sensory analyses of food products by hyperspectral imaging: Recent application developments" in Trends in Food Science & Technology (IF = 11.
077) In summary, the application progress of hyperspectral imaging in fast and non-destructive sensory analysis of food is systematically described.
Academician Sun Dawen of South China University of Technology is the only corresponding author of this article.

background

Sensory analysis is the evaluation of signals received by vision, hearing, taste, smell and touch.
Traditional methods of analyzing sensory characteristics including trained sensory panels, colorimeters and texture analyzers are invasive, laborious and small-scale procedures.
Hyperspectral imaging system (HSI) combines imaging and spectroscopy to obtain spatial and spectral data in one system.
It has become a more time-saving and non-destructive method for determining the sensory properties of various foods .

Summary content

This review aims to provide comprehensive information on the application of hyperspectral imaging technology in the evaluation of food sensory characteristics since 2010.
This article briefly introduces the hyperspectral imaging technology, discusses the key steps of image processing and the application progress in the evaluation of color, defect, texture, flavor, freshness and maturity.
In addition, this article also introduces the advantages and challenges of HSI technology, and looks forward to future research directions.

in conclusion

Vis-NIR (400-1000 nm) hyperspectral imaging is mainly used for the evaluation of sensory characteristics.
In addition, in the sensory evaluation of HSI, the commonly used multivariate analysis is the linear regression algorithm (PLSR and MLR), while the nonlinear analysis performs better in sensory feature prediction.
The regression model is used to determine the texture, color, flavor and freshness of the food, and the classification model is used to detect defects and maturity to obtain the most accurate results.
In addition, the combination of algorithms and the fusion of spectral and spatial information have obtained more accurate results in the evaluation of sensory features.
Although the improvement of the HSI system still has certain limitations, the application of HSI to the real-time monitoring of sensory performance has proven to have great potential in industrial applications.

Although hyperspectral imaging technology has shown good performance in predicting the sensory evaluation of agricultural products, there are still some challenges that need to be considered in future research: For example, in sensory evaluation research, the most commonly used image acquisition is the reflectance mode.
However, it is a challenge to use this model to accurately analyze round food products such as fruits.
In recent years, the scattering mode has also been applied and has shown higher accuracy than the reflection mode.
However, the application of other image acquisition modes in sensory evaluation is very limited, so these modes need to be studied in future research; HSI is in Approximately 70% of sensory evaluation applications are performed in the 400-1000 nm spectral range, while only about 25% are performed in the 900-1700 nm spectral range.
However, the application of HSI in the spectral regions of UV (220-400 nm), SWIR (1000-2500 nm) and MWIR (3500-5000 nm) is very limited, so the performance of these spectral ranges is still a mystery.
Therefore, future research can explore the feasibility of these spectral ranges to further advance this technology; the application of hyperspectral imaging in sensory evaluation is mostly limited to the laboratory.
Before the industrial-scale application of HSI, the prototype can be used for sensory evaluation of some modified equipment, such as the use of micro-stages to adjust the sample distance based on samples of different sizes and shapes, the use of thermoelectric cooling systems to reduce system noise, and increase the operating temperature.
A very long conveyor belt simulates real-time online image capture and so on.

Attached

Expert profile

Professor Da-Wen Sun, a member of the Central Committee of China Zhi Gong Party, a special expert of the Chinese Federation of Returned Overseas Chinese, and a member of the Expert Advisory Committee of the Overseas Chinese Affairs Office of the State Council.
Academician of the European Academy of Sciences (European Academy of Humanities and Natural Sciences), academician of the Royal Irish Academy of Sciences, academician of the Polish Academy of Sciences, academician of the International Academy of Food Sciences, academician of the International Academy of Agriculture and Biosystems Engineering, academician of the International Academy of Refrigeration Sciences.

Won the International Council of Agriculture and Biological Systems Engineering (CIGR) Outstanding Award, the Royal Institution of Mechanical Engineers "Food Engineer of the Year", Phoenix Satellite TV "Influencing the World Chinese Award", International Food Protection Association "Frozen Food Foundation Frozen Research Award", International Engineering and Food Association Lifetime Achievement Award, "CIGR Honorary Chairman" and many other international awards, from 2015 to 2020, won the "Highly Cited Scientist" title of Clarivate Safety Ball for six consecutive years.

Published 1,000 papers in world-renowned magazines and international conferences, published 17 monographs, and a total of more than 600 papers were included in SCI (Web of Science's academic h index is 104, SCOPUS's academic h index is 108, GoogleScholar statistics The academic h index is 122).

Since returning to work at South China University of Technology at the end of 2011, among the articles that can be retrieved on the Web ofScience website so far (November 2020), a total of 209 articles are based on the School of Food Science and Engineering of South China University of Technology as the first author unit.

Authorized 63 invention patents in China, 10 invention patents in the United States, Britain and Japan, and applied for 9 international PCT patents.

The article "Rapid and noninvasive sensory analyses of food products by hyperspectral imaging: Recent application developments" was published online in Trends in Food Science & Technology .