一种基于双通道模型的土壤有机质含量预测方法

By combining temporal and spatial feature extraction with a dual-channel model, the problem of unutilized band correlation in spectral data was solved, achieving higher accuracy in predicting soil organic matter content.

CN119989442BActive Publication Date: 2026-07-17HEILONGJIANG UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HEILONGJIANG UNIV
Filing Date
2025-01-15
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing technologies fail to fully utilize the potential connections between multiple bands in spectral data, resulting in insufficient accuracy in predicting soil organic matter content.

Method used

A dual-channel model-based approach is adopted, which models spectral data and climate and terrain data through temporal feature extraction channels and spatial feature extraction channels, respectively. The Huber loss function is used to optimize the model, and the output features of the two channels are fused for prediction.

Benefits of technology

It improves the accuracy of soil organic matter content prediction, enhances the model's ability to detect outliers, and significantly surpasses the prediction performance of existing models.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure QLYQS_9
    Figure QLYQS_9
  • Figure QLYQS_15
    Figure QLYQS_15
  • Figure QLYQS_20
    Figure QLYQS_20
Patent Text Reader

Abstract

一种基于双通道模型的土壤有机质含量预测方法，涉及基于深度学习模型的土壤有机质制图领域。本发明是为了解决土壤有机质领域现有建模方法未能充分利用光谱数据中多个波段间的潜在联系的问题。本发明包括：获取待预测区域的光谱数据和气候与地形环境协变量数据，将上述两类数据输入到训练好的2C‑Net模型中，获得土壤有机质含量预测结果；2C‑Net模型包括：时间特征提取通道、空间特征提取通道、预测头；时间特征提取通道用于捕获光谱数据中每个波段自身与多个波段之间在时间维度上的特征，得到时间通道输出特征；空间特征提取通道用于对气候和地形环境协变量数据进行空间维度的建模，得到空间通道输出特征；预测头将时间通道输出特征与空间通道输出特征融合，进行土壤有机质含量预测。本发明用于实现土壤有机质含量预测。
Need to check novelty before this filing date? Find Prior Art