Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis
With the rapid progression of agricultural informatization technology, the methodologies of crop monitoring based on spectral technology are constantly upgraded. In order to carry out the efficient, precise and nondestructive detection of relative chlorophyll (SPAD) during the booting stage, we acqu...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/5/1693 |
| _version_ | 1797263811232661504 |
|---|---|
| author | Hongwei Cui Haolei Zhang Hao Ma Jiangtao Ji |
| author_facet | Hongwei Cui Haolei Zhang Hao Ma Jiangtao Ji |
| author_sort | Hongwei Cui |
| collection | DOAJ |
| description | With the rapid progression of agricultural informatization technology, the methodologies of crop monitoring based on spectral technology are constantly upgraded. In order to carry out the efficient, precise and nondestructive detection of relative chlorophyll (SPAD) during the booting stage, we acquired hyperspectral reflectance data about spring wheat vertical distribution and adopted the fractional-order differential to transform the raw spectral data. After that, based on correlation analysis, fractional differential spectra and fractional differential spectral indices with strong correlation with SPAD were screened and fused. Then, the least-squares support vector machine (LSSSVM) and the least-squares support vector machine (SMA-LSSSVM) optimized on the slime mold algorithm were applied to construct the estimation models of SPAD, and the model accuracy was assessed to screen the optimal estimation models. The results showed that the 0.4 order fractional-order differential spectra had the highest correlation with SPAD, which was 9.3% higher than the maximum correlation coefficient of the original spectra; the constructed two-band differential spectral indices were more sensitive to SPAD than the single differential spectra, in which the correlation reached the highest level of 0.724. The SMA-LSSSVM model constructed based on the two-band fractional-order differential spectral indices was better than the single differential spectra and the integration of both, which realized the assessment of wheat SPAD. |
| first_indexed | 2024-04-25T00:18:56Z |
| format | Article |
| id | doaj.art-7c1d20fe1fce4fd78623d55397be7b6d |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-25T00:18:56Z |
| publishDate | 2024-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-7c1d20fe1fce4fd78623d55397be7b6d2024-03-12T16:55:39ZengMDPI AGSensors1424-82202024-03-01245169310.3390/s24051693Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral AnalysisHongwei Cui0Haolei Zhang1Hao Ma2Jiangtao Ji3College of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471000, ChinaCollege of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471000, ChinaCollege of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471000, ChinaCollege of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471000, ChinaWith the rapid progression of agricultural informatization technology, the methodologies of crop monitoring based on spectral technology are constantly upgraded. In order to carry out the efficient, precise and nondestructive detection of relative chlorophyll (SPAD) during the booting stage, we acquired hyperspectral reflectance data about spring wheat vertical distribution and adopted the fractional-order differential to transform the raw spectral data. After that, based on correlation analysis, fractional differential spectra and fractional differential spectral indices with strong correlation with SPAD were screened and fused. Then, the least-squares support vector machine (LSSSVM) and the least-squares support vector machine (SMA-LSSSVM) optimized on the slime mold algorithm were applied to construct the estimation models of SPAD, and the model accuracy was assessed to screen the optimal estimation models. The results showed that the 0.4 order fractional-order differential spectra had the highest correlation with SPAD, which was 9.3% higher than the maximum correlation coefficient of the original spectra; the constructed two-band differential spectral indices were more sensitive to SPAD than the single differential spectra, in which the correlation reached the highest level of 0.724. The SMA-LSSSVM model constructed based on the two-band fractional-order differential spectral indices was better than the single differential spectra and the integration of both, which realized the assessment of wheat SPAD.https://www.mdpi.com/1424-8220/24/5/1693spring wheatbooting stageSPADfractional-order differentiationSMA-LSSSVMfractional-order differentiation spectral index |
| spellingShingle | Hongwei Cui Haolei Zhang Hao Ma Jiangtao Ji Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis Sensors spring wheat booting stage SPAD fractional-order differentiation SMA-LSSSVM fractional-order differentiation spectral index |
| title | Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis |
| title_full | Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis |
| title_fullStr | Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis |
| title_full_unstemmed | Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis |
| title_short | Research on SPAD Estimation Model for Spring Wheat Booting Stage Based on Hyperspectral Analysis |
| title_sort | research on spad estimation model for spring wheat booting stage based on hyperspectral analysis |
| topic | spring wheat booting stage SPAD fractional-order differentiation SMA-LSSSVM fractional-order differentiation spectral index |
| url | https://www.mdpi.com/1424-8220/24/5/1693 |
| work_keys_str_mv | AT hongweicui researchonspadestimationmodelforspringwheatbootingstagebasedonhyperspectralanalysis AT haoleizhang researchonspadestimationmodelforspringwheatbootingstagebasedonhyperspectralanalysis AT haoma researchonspadestimationmodelforspringwheatbootingstagebasedonhyperspectralanalysis AT jiangtaoji researchonspadestimationmodelforspringwheatbootingstagebasedonhyperspectralanalysis |