FT-NIR Spectrometer,fourier transform spectrometer,ftir spectrometer,ftir analysis,Fourier Transform Infrared Spectroscopy,spectral analysis,quantitative analysis,qualitative analysis,diffuse reflection,specular reflection,integrating sphere,FT-IR Spectrometer-Advanced Manufacturer,FT-NIR Spectrometers；ftir spectrometer,Optosky is a manufacturer of advanced FT-IR spectrometers, with products in stock and at affordable prices.,lntegrating Sphere,integrating sphere,specular reflection,diffuse reflection,qualitative analysis,quantitative analysis,spectral analysis,Fourier Transform Infrared Spectroscopy,ftir analysis,ftir spectrometer,fourier transform spectrometer,FT-IR Spectrometer-Advanced Manufacturer,FT-NIR Spectrometers；ftir spectrometer,Optosky is a manufacturer of advanced FT-IR spectrometers, with products in stock and at affordable prices.

Estimation of Forage Biomass and Nutritional Value by Ground Spectrometer

author: Gavin

2022-03-08

Accurate estimation of pasture characteristics is an important basis for guiding farmers to formulate fertilization and harvesting plans. The most common method for estimating the nutritional value of pastures is laboratory analysis, which limits decision-making in the field. The ground object spectrometer can obtain crop biochemical information non-destructively by measuring and analyzing the canopy spectral reflectance (CSR). And crop height is an important agronomic parameter closely related to forage quality. Canopy spectral reflectance combined with crop height can improve estimates of cellulose and protein content in heterogeneous pastures compared to spectral information alone, and is an effective tool for assessing forage yield and quality in the field.

Principle:
Estimation of forage biomass and nutritional value using PLSR and SVM in combination with canopy spectroscopy and auxiliary growth parameters such as crop height.

Solution:
The yield and quality parameters of the timothy grass and clover mixture were obtained from field experiments, and their canopy spectral reflectance values (CSR) were measured by ATP9110-25, and then analyzed by partial least squares regression (PLSR) and support vector machine (SVM). CSR data and crop height information to estimate forage biomass value and nutritional value, with particular attention to whether including timothy grass and clover height as model input variables would improve prediction performance.

Test result:

Verify that SVMspec and SVMspec+H are used for estimation

(a) dry matter yield,
(b) in vitro true digestibility,
(c) neutral detergent fibers,
(d) neutral detergent fiber digestibility,
(e) acid detergent fibers,
(f) acid washed lignin,
(g) crude protein,
(h) Crude protein yield
(i) Plant composition.
The black solid and dashed lines represent the linear regression fit lines of PLSRspec+H and PLSRspec, respectively.

Verify that SVMspec and SVMspec+H are used for estimation
(a) dry matter yield,
(b) in vitro true digestibility,
(c) neutral detergent fibers,
(d) neutral detergent fiber digestibility,
(e) acid detergent fibers,
(f) acid washed lignin,
(g) crude protein,
(h) Crude protein yield
(i) Plant composition.
The black solid and dashed lines represent the linear regression fit lines of SVMspec+H and SVMspec, respectively.

Using the PLSRspec+H model to
(a) dry matter yield,
(b) in vitro true digestibility,
(c) neutral detergent fibers,
(d) neutral detergent fiber digestibility,
(e) acid detergent fibers,
(f) acid washed lignin,
(g) crude protein,
(h) Crude protein yield
(i) Plant composition
Variable Projected Importance (VIP) values for spectral variables.

Conclution:
The study demonstrates the feasibility of estimating forage biomass and nutritional value using PLSR and SVM in combination with canopy spectroscopy and auxiliary growth parameters such as crop height. In terms of R2 and RRMSE, SVM performs well, especially when estimating NDF and CPY, although there is significant overfitting when estimating some variables such as IVTD and ADL. In conclusion, PLSR exhibits stable prediction results between calibration and validation sets. Furthermore, the PLSRspec+H model utilizing both explanatory variables outperformed the PLSRspec model using only the full spectrum. PLSRwave+H has great potential to simplify estimation models by utilizing fewer explanatory variables.

Reconmand product: