Rapid assessment of coniferous biomass lignin–carbohydrates with near-infrared spectroscopy

Effective and rapid assessment of pork freshness is significant for monitoring pork quality. However, a traditional sensory evaluation method is subjective and physicochemical analysis is time-consuming. In this study, the near-infrared spectroscopy (NIRS) technique, a fast and non-destructive analysis method, is employed to determine pork freshness. Considering that commonly used statistical modeling methods require preprocessing data for satisfactory performance, this paper presents a one-dimensional squeeze-and-excitation residual network (1D-SE-ResNet) to construct the complex relationship between pork freshness and NIRS. The developed model enhances the one-dimensional residual network (1D-ResNet) with squeeze-and-excitation (SE) blocks. As a deep learning model, the proposed method is capable of extracting features from the input spectra automatically and can be used as an end-to-end model to simplify the modeling process. A comparison between the proposed method and five popular classification models indicates that the 1D-SE-ResNet achieves the best performance, with a classification accuracy of 93.72%. The research demonstrates that the NIRS analysis technique based on deep learning provides a promising tool for pork freshness detection and therefore is helpful for ensuring food safety.

Download Full-text

Rapid Assessment of Fresh Beef Spoilage Using Portable Near-Infrared Spectroscopy

Jurnal Keteknikan Pertanian ◽

10.19028/jtep.09.3.79-86 ◽

2021 ◽

Vol 9 (3) ◽

pp. 79-86

Author(s):

Akeme Cyril Njume ◽

Y. Aris Purwanto ◽

Dewi Apri Astuti ◽

Slamet Widodo

Keyword(s):

Infrared Spectroscopy ◽

Near Infrared Spectroscopy ◽

Near Infrared ◽

Rapid Assessment ◽

Partial Least Square ◽

Least Square ◽

Partial Least Square Regression ◽

Robust Model ◽

Acceptable Model ◽

Assess Quality

The objective of this study was to develop a prediction model to assess fresh beef spoilage directly with the use of a portable near-infrared spectroscopy (NIRS), without conducting a chemical method. Three fresh beef samples were bought from a slaughterhouse and traditional market on separate days. Spectra were acquired using a portable Scio spectrometer with wavelength 740-1070 nm, and two-third was used for calibration sets and one-third for validation sets. Partial least square regression and cross-validation were used to develop a model and equation for predicting beef spoilage. The changes observed were changed in color, water loss, and muscle hardness. The best predictive model was obtained from the original spectra (no pre-process) results as follows (R2C = 0.9, Rp = 0.86, SEC = 0.61, SEP = 0.69 and RPD = 3.53). Multiple Scattered Correlation (MSC) pre-processing method gave a good and acceptable model with results as follows; Rc = 0.89, SEC = 0.66, SEP = 0.83 and RPD = 2.91. NIRS showed variability of the samples and rate of spoilage, hence, can be used to assess quality and safety. Further studies are needed to develop a robust model to predict fresh beef spoilage using a portable NIRS Scio.

Download Full-text

Non-Destructive Prediction of the Properties of Forest Biomass for Chemical and Bioenergy Applications Using near Infrared Spectroscopy

Journal of Near Infrared Spectroscopy ◽

10.1255/jnirs.1153 ◽

2015 ◽

Vol 23 (2) ◽

pp. 93-102 ◽

Cited By ~ 16

Author(s):

Gifty E. Acquah ◽

Brian K. Via ◽

Oladiran O. Fasina ◽

Lori G. Eckhardt

Keyword(s):

Infrared Spectroscopy ◽

Near Infrared Spectroscopy ◽

Loblolly Pine ◽

Near Infrared ◽

Rapid Assessment ◽

Forest Biomass ◽

Volatile Matter ◽

Least Squares Regression ◽

Distribution Models ◽

Single Spectrum

Forest biomass will play a key role as a feedstock for bioproducts as the bioeconomy develops. Rapid assessment of this heterogeneous resource will help determine its suitability as feedstock for specific applications, aid in feedstock improvement programmes and enable better process control that will optimise the biorefinery process. In this study, near infrared spectroscopy coupled with partial least-squares regression was used to predict important chemical and thermal reactivity properties of biomass made up of needles, twigs, branches, bark and wood of Pinus taeda (loblolly pine). Models developed with the raw spectra for property prediction used between three and eight factors to yield R2 values ranging from a low of 0.34 for higher heat values to a high of 0.92 for volatile matter. Pretreating the raw spectra with first derivatives improved the fit statistics for all properties (i.e. min 0.57, max 0.92; with two or three factors). The best-performing models were for extractives, lignin, glucose, cellulose, volatile matter and fixed carbon ( R2 ≥ 0.80, residual predictive deviation/ratio of performance to deviation ≥1.5). This study provided the capacity to predict multiple chemical and thermal/energy traits from a single spectrum across an array of materials that differ considerably in chemistry type and distribution. Models developed should be able to rapidly predict the studied properties of similar biomass types. This will be useful in rapidly allocating feedstocks that optimise biomass conversion technologies.

Download Full-text