Non-destructive detection of dicyandiamide in infant formula powder using multi-spectral imaging coupled with chemometrics

The rapid and non-destructive detection of mechanical damage to fruit during postharvest supply chains is important for monitoring fruit deterioration in time and optimizing freshness preservation and packaging strategies. As fruit is usually packed during supply chain operations, it is difficult to detect whether it has suffered mechanical damage by visual observation and spectral imaging technologies. In this study, based on the volatile substances (VOCs) in yellow peaches, the electronic nose (e-nose) technology was applied to non-destructively predict the levels of compression damage in yellow peaches, discriminate the damaged fruit and predict the time after the damage. A comparison of the models, established based on the samples at different times after damage, was also carried out. The results show that, at 24 h after damage, the correct answer rate for identifying the damaged fruit was 93.33%, and the residual predictive deviation in predicting the levels of compression damage and the time after the damage, was 2.139 and 2.114, respectively. The results of e-nose and gas chromatography-mass spectrophotometry (GC–MS) showed that the VOCs changed after being compressed—this was the basis of the e-nose detection. Therefore, the e-nose is a promising candidate for the detection of compression damage in yellow peach.

Download Full-text

Detection of Adulteration in Infant Formula Based on Ensemble Convolutional Neural Network and Near-Infrared Spectroscopy

Foods ◽

10.3390/foods10040785 ◽

2021 ◽

Vol 10 (4) ◽

pp. 785

Author(s):

Yisen Liu ◽

Songbin Zhou ◽

Wei Han ◽

Chang Li ◽

Weixin Liu ◽

...

Keyword(s):

Neural Network ◽

Convolutional Neural Network ◽

Infant Formula ◽

Ensemble Learning ◽

Near Infrared ◽

Nir Spectroscopy ◽

Data Sets ◽

Promising Tool ◽

Random Initialization ◽

Non Destructive

Adulteration in dairy products has received world-wide attention, and at the same time, near infrared (NIR) spectroscopy has proven to be a promising tool for adulteration detection given its advantages of real-time response and non-destructive analysis. Regardless, the accurate and robust NIR model for adulteration detection is hard to achieve in practice. Convolutional neural network (CNN), as a promising deep learning architecture, is difficult to apply to such chemometrics tasks due to the high risk of overfitting, despite the breakthroughs it has made in other fields. In this paper, the ensemble learning method based on CNN estimators was developed to address the overfitting and random initialization problems of CNN and applied to the determination of two infant formula adulterants, namely hydrolyzed leather protein (HLP) and melamine. Moreover, a probabilistic wavelength selection method based on the attention mechanism was proposed for the purpose of finding the best trade-off between the accuracy and the diversity of the sub-models in ensemble learning. The overall results demonstrate that the proposed method yielded superiority regression performance over the comparison methods for both studied data sets, and determination coefficients (R2) of 0.961 and 0.995 were obtained for the HLP and the melamine data sets, respectively.

Download Full-text

Quantitative and non-destructive evaluation of ground beef based on multi-spectral imaging

2020 IEEE International Conference on Industrial Technology (ICIT) ◽

10.1109/icit45562.2020.9067300 ◽

2020 ◽

Author(s):

Omar Gutierrez-Navarro ◽

Daniel U. Campos-Delgado ◽

Rafael A. Casillas Penuelas ◽

Carlos U Haubi Segura

Keyword(s):

Ground Beef ◽

Spectral Imaging ◽

Non Destructive Evaluation ◽

Non Destructive

Download Full-text

Autofluorescence-spectral imaging as an innovative method for rapid, non-destructive and reliable assessing of soybean seed quality

Scientific Reports ◽

10.1038/s41598-021-97223-5 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Clíssia Barboza da Silva ◽

Nielsen Moreira Oliveira ◽

Marcia Eugenia Amaral de Carvalho ◽

André Dantas de Medeiros ◽

Marina de Lima Nogueira ◽

...

Keyword(s):

Machine Learning ◽

High Performance ◽

Seed Quality ◽

Soybean Seed ◽

Spectral Imaging ◽

Machine Learning Algorithms ◽

Support Vector ◽

Soybean Seeds ◽

Linear Discriminant ◽

Non Destructive

AbstractIn the agricultural industry, advances in optical imaging technologies based on rapid and non-destructive approaches have contributed to increase food production for the growing population. The present study employed autofluorescence-spectral imaging and machine learning algorithms to develop distinct models for classification of soybean seeds differing in physiological quality after artificial aging. Autofluorescence signals from the 365/400 nm excitation-emission combination (that exhibited a perfect correlation with the total phenols in the embryo) were efficiently able to segregate treatments. Furthermore, it was also possible to demonstrate a strong correlation between autofluorescence-spectral data and several quality indicators, such as early germination and seed tolerance to stressful conditions. The machine learning models developed based on artificial neural network, support vector machine or linear discriminant analysis showed high performance (0.99 accuracy) for classifying seeds with different quality levels. Taken together, our study shows that the physiological potential of soybean seeds is reduced accompanied by changes in the concentration and, probably in the structure of autofluorescent compounds. In addition, altering the autofluorescent properties in seeds impact the photosynthesis apparatus in seedlings. From the practical point of view, autofluorescence-based imaging can be used to check modifications in the optical properties of soybean seed tissues and to consistently discriminate high-and low-vigor seeds.

Download Full-text