Crop canopy light field spectrogram data correction method and nutrient detection device

By constructing a weighted fusion strategy of a light field heterogeneity characterization parameter model and a dual regression model, the problems of spectral distortion and insufficient nutrient inversion accuracy of crop canopy hyperspectral images under non-uniform field illumination were solved, and adaptive correction of spectral data and accurate nutrient detection were achieved.

CN122265612APending Publication Date: 2026-06-23CHINA AGRI UNIV
0 Cites 0 Cited by

Patent Information

Application Number
CN202610305869.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-13
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

Under non-uniform lighting conditions in the field, existing technologies cannot effectively handle the uneven lighting phenomenon in hyperspectral images, resulting in crop canopy data distortion and insufficient nutrient inversion accuracy. Traditional spectral correction methods ignore the differences between strong light areas and shadow areas, leading to loss of texture details or noise residue, and single regression models are difficult to adapt to dynamic light environments.

Method used

A model for characterizing the heterogeneity of the light field, including the red-edge modulation coefficient, is constructed. Strong and weak light regions are identified by clustering. A benchmark correction model is constructed and weighted fusion is performed using fusion weight coefficients. Nutrient content is calculated by combining a bi-regression model, thereby achieving adaptive correction and accurate inversion of the spectrum.

Benefits of technology

It improves the accuracy of crop canopy nutrient detection and the authenticity of spatial distribution maps, solves the problem of spectral distortion under uneven lighting conditions, achieves a smooth spectral transition from strong light areas to shaded areas, and ensures the spatial continuity and detection accuracy of the corrected spectral data.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to the technical field of spectroscopy detection, and discloses a crop canopy light field spectrogram data correction method and a nutrient detection device.The method comprises the steps that crop canopy hyperspectral data are collected, and after radiation correction, vegetation indexes are adjusted through optimized soil to remove backgrounds; a light field heterogeneity parameter model combined with a red edge response factor is constructed, and a red edge modulation coefficient is determined through clustering evaluation index iterative optimization; calculating the Euclidean distance from the pixel to the strong and weak light clustering center to generate a fusion weight coefficient; and constructing a differential benchmark correction model, and performing two times of linear weighted fusion on the spectral data and the nutrient content calculated by the double-regression model by using the fusion weight. By constructing light field heterogeneity characterization parameters and a distance weight-based soft segmentation strategy, smooth transition of strong and weak light areas is realized, spectrum distortion caused by uneven illumination is eliminated, and the accuracy and robustness of crop nutrient detection in a complex light field are improved.
Need to check novelty before this filing date? Find Prior Art