基于无人机多光谱成像的补充耕地土壤肥力快速评估装置与方法
By using UAV multispectral imaging to analyze soil spectral variations in real time, dynamically generating sampling points and constructing a calibration dataset, the problem of low accuracy in soil fertility assessment models for supplementary cultivated land is solved, and efficient and accurate soil fertility assessment is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGXI ZHUANG AUTONOMOUS REGION NATURAL RESOURCES ECOLOGICAL RESTORATION CENT
- Filing Date
- 2026-05-06
- Publication Date
- 2026-07-17
AI Technical Summary
In the replenishment of arable land, due to the complex soil sources and high spatial heterogeneity, traditional soil fertility assessment methods based on fixed sampling points are difficult to effectively capture the spatial variation of soil fertility, resulting in low model calibration accuracy and affecting the inversion accuracy of key indicators such as soil organic matter and available nitrogen.
By analyzing soil spectral variations in real time using UAV multispectral imaging, additional sampling points are dynamically generated. A calibration dataset is constructed by combining the base sampling points, and a localized calibration model is used for soil inversion, thereby improving the model's adaptability and accuracy.
It improves the accuracy of soil fertility parameter inversion and the robustness of the model, enabling rapid and accurate assessment of soil fertility in replenished arable land, and reducing labor and time costs.