Spectra transfer based learning for predicting and classifying soil texture with short-ranged Vis-NIRS sensor

Munnaf, Muhammad Abdul; Mouazen, Abdul Mounem

doi:10.1016/j.still.2022.105545

Cited by 13 publications

(6 citation statements)

References 65 publications

(68 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other authors presented proposals similar to the approach used in this paper based on light spectra analyses. In [ 38 ], the authors used visible (Vis) and near-infrared (NIR) data, while in [ 39 ], they also used pXRF information. In both cases, accuracies were below 100%, ranging from 70% to 90% with Vis and NIR data and 97% when X pXRF-ray information was added.…”

Section: Discussionmentioning

confidence: 99%

“…In our case, the class corresponded to the percentage of sand and clay in the sample. In [ 38 , 40 , 41 , 42 ], the authors also used the percentage of soil components as a predictor.…”

Section: Discussionmentioning

confidence: 99%

“…Among different data-processing techniques in the included studies, there is a wide variety of methods, such as convolutional neural network (CNN) [ 27 ], RF [ 39 ], partial least squares regression (PLSR) [ 38 , 40 , 41 ], and multivariate image analysis (MIA) [ 24 ]. In some cases, even two different methods are combined, such as recursive feature elimination (RFE) with RF [ 42 ].…”

Section: Discussionmentioning

confidence: 99%

See 2 more Smart Citations

Low-Cost Optical Sensors for Soil Composition Monitoring

Diaz,

Ahmad,

Parra

et al. 2024

Sensors

View full text Add to dashboard Cite

Studying soil composition is vital for agricultural and edaphology disciplines. Presently, colorimetry serves as a prevalent method for the on-site visual examination of soil characteristics. However, this technique necessitates the laboratory-based analysis of extracted soil fragments by skilled personnel, leading to substantial time and resource consumption. Contrastingly, sensor techniques effectively gather environmental data, though they mostly lack in situ studies. Despite this, sensors offer substantial on-site data generation potential in a non-invasive manner and can be included in wireless sensor networks. Therefore, the aim of the paper is to develop a low-cost red, green, and blue (RGB)-based sensor system capable of detecting changes in the composition of the soil. The proposed sensor system was found to be effective when the sample materials, including salt, sand, and nitro phosphate, were determined under eight different RGB lights. Statistical analyses showed that each material could be classified with significant differences based on specific light variations. The results from a discriminant analysis documented the 100% prediction accuracy of the system. In order to use the minimum number of colors, all the possible color combinations were evaluated. Consequently, a combination of six colors for salt and nitro phosphate successfully classified the materials, whereas all the eight colors were found to be effective for classifying sand samples. The proposed low-cost RGB sensor system provides an economically viable and easily accessible solution for soil classification.

show abstract

Section: Discussionmentioning

confidence: 99%

“…In our case, the class corresponded to the percentage of sand and clay in the sample. In [ 38 , 40 , 41 , 42 ], the authors also used the percentage of soil components as a predictor.…”

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Low-Cost Optical Sensors for Soil Composition Monitoring

Diaz,

Ahmad,

Parra

et al. 2024

Sensors

View full text Add to dashboard Cite

show abstract

“…A 50 W light source using a fiber optic cable was used to illuminate soil samples. This fiber optic cable (Analytical Spectral Devices, Inc., USA) was deployed with a confocal lens fitted in a lens holder (Figure 2) at 18º within the lens (Munnaf and Mouazen 2023). As mentioned above, soil samples were taken out of the incubation tubes at different days for mineral N analysis, and vis-NIR scanning.…”

Section: Laboratory Vis-nirs Scanningmentioning

confidence: 99%

Visible and near infrared spectroscopy for predicting soil nitrogen mineralization rate: Effect of incubation period and ancillary soil properties

Ruma,

Munnaf,

De Neve

et al. 2024

CATENA

View full text Add to dashboard Cite

“…Typical model transfer algorithms include Direct Standardization (DS) [20], Piecewise Direct Standardization (PDS), Double Window Piecewise Direct Standardization (DW-PDS) [21], and Orthogonal Signal Correction (OSC) [22]. These algorithms could also be applied to transfer models between different physical states of the same substance [23][24][25]. Pereira et al [23] applied the DWPDS method to model the delivery of drug powder mixtures to complete tablet samples.…”

Section: Introductionmentioning

confidence: 99%

Rapid Detection of Total Viable Count in Intact Beef Dishes Based on NIR Hyperspectral Hybrid Model

Wei,

Zhang,

et al. 2023

Sensors

View full text Add to dashboard Cite

The total viable count (TVC) of bacteria is an important index to evaluate the freshness and safety of dishes. To improve the accuracy and robustness of spectroscopic detection of total viable bacteria count in a complex system, a new method based on a near-infrared (NIR) hyperspectral hybrid model and Support Vector Machine (SVM) algorithms was developed to directly determine the total viable count in intact beef dish samples in this study. Diffuse reflectance data of intact and crushed samples were tested by NIR hyperspectral and processed using Multiplicative Scattering Correction (MSC) and Competitive Adaptive Reweighted Sampling (CARS). Kennard–Stone (KS) and Samples Set Partitioning Based on Joint X-Y Distance (SPXY) algorithms were used to select the optimal number of standard samples transferred by the model combined with root mean square error. The crushed samples were transferred into the complete samples prediction model through the Direct Standardization (DS) algorithm. The spectral hybrid model of crushed samples and full samples was established. The results showed that the Determination Coefficient of Calibration (RP2) value of the total samples prediction set increased from 0.5088 to 0.8068, and the value of the Root Mean Square Error of Prediction (RMSEP) decreased from 0.2454 to 0.1691 log10 CFU/g. After establishing the hybrid model, the RMSEP value decreased by 9.23% more than before, and the values of Relative Percent Deviation (RPD) and Reaction Error Relation (RER) increased by 12.12% and 10.09, respectively. The results of this study showed that TVC instewed beef samples can be non-destructively determined based on the DS model transfer method combined with the hybrid model strategy. This study provided a reference for solving the problem of poor accuracy and reliability of prediction models in heterogeneous samples.

show abstract

Spectra transfer based learning for predicting and classifying soil texture with short-ranged Vis-NIRS sensor

Cited by 13 publications

References 65 publications

Low-Cost Optical Sensors for Soil Composition Monitoring

Low-Cost Optical Sensors for Soil Composition Monitoring

Visible and near infrared spectroscopy for predicting soil nitrogen mineralization rate: Effect of incubation period and ancillary soil properties

Rapid Detection of Total Viable Count in Intact Beef Dishes Based on NIR Hyperspectral Hybrid Model

Contact Info

Product

Resources

About