Soil Biology and Biochemistry, Journal Year: 2024, Volume and Issue: unknown, P. 109699 - 109699
Published: Dec. 1, 2024
Language: Английский
Soil Biology and Biochemistry, Journal Year: 2024, Volume and Issue: unknown, P. 109699 - 109699
Published: Dec. 1, 2024
Language: Английский
Journal of soil science and plant nutrition, Journal Year: 2025, Volume and Issue: unknown
Published: Jan. 31, 2025
Language: Английский
Citations
0Applied Soil Ecology, Journal Year: 2025, Volume and Issue: 207, P. 105922 - 105922
Published: Feb. 11, 2025
Language: Английский
Citations
0Journal of Hazardous Materials, Journal Year: 2025, Volume and Issue: unknown, P. 138162 - 138162
Published: April 1, 2025
Language: Английский
Citations
0Agricultural Water Management, Journal Year: 2025, Volume and Issue: 313, P. 109473 - 109473
Published: April 9, 2025
Language: Английский
Citations
0Environmental Technology & Innovation, Journal Year: 2025, Volume and Issue: 38, P. 104070 - 104070
Published: Feb. 7, 2025
Language: Английский
Citations
0Environmental Technology & Innovation, Journal Year: 2024, Volume and Issue: 36, P. 103848 - 103848
Published: Oct. 6, 2024
Language: Английский
Citations
2Agriculture, Journal Year: 2024, Volume and Issue: 14(7), P. 1186 - 1186
Published: July 18, 2024
As a vital pigment for photosynthesis in rice, chlorophyll content is closely correlated with growth status and photosynthetic capacity. The estimation of allows the monitoring rice facilitates precise management field, such as application fertilizers irrigation. advancement hyperspectral remote sensing technology has made it possible to estimate non-destructively, quickly, effectively, offering technical support managing across wide areas. Although data have fine spectral resolution, they also cause large amount information redundancy noise. This study focuses on issues unstable input variables model’s poor applicability various periods when predicting content. By introducing theory harmonic analysis time-frequency conversion method, deep neural network (DNN) model framework based wavelet packet transform-first order differential-harmonic (WPT-FD-HA) was proposed, which avoids uncertainty calculation parameters. accuracy estimating WPT-FD WPT-FD-HA compared at seedling, tillering, jointing, heading, grain filling, milk, complete evaluate validity generalizability suggested framework. results demonstrated that all models’ single-period validation had coefficients determination (R2) values greater than 0.9 RMSE less 1. multi-period root mean square error (RMSE) 1.664 an R2 0.971. Even independent splitting validation, can still achieve = 0.95 1.4. WPT-FD-HA-based learning exhibited strong stability. outcome this deserves be used monitor broad scale using data.
Language: Английский
Citations
1Soil Biology and Biochemistry, Journal Year: 2024, Volume and Issue: unknown, P. 109699 - 109699
Published: Dec. 1, 2024
Language: Английский
Citations
1