Sea ice concentration retrieval method and device based on brightness temperature data and electronic equipment

By employing multi-scale feature extraction and cross-feature concatenation methods, combined with dense residual coding, MSPC-ResNet is constructed to address the issues of atmospheric interference and nonlinear variations in sea ice concentration inversion, achieving higher inversion accuracy and stability.

CN122334055APending Publication Date: 2026-07-03CENT SOUTH UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610808941.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-05
Publication Date
2026-07-03

Smart Images

  • Figure CN122334055A_ABST
    Figure CN122334055A_ABST
Patent Text Reader

Abstract

This invention discloses a method, apparatus, and electronic device for sea ice concentration inversion based on brightness temperature data, belonging to the field of sea ice concentration inversion technology. The method includes: inputting brightness temperature observation vectors into multiple scale branches; extracting features from the brightness temperature observation vectors at different scales to obtain multiple scale features; generating multiple feature factors from the brightness temperature observation vectors through multiple sets of low-dimensional mappings; and performing element-wise multiplication on the multiple feature factors to obtain cross features; concatenating the multiple scale features and the cross features to obtain an initial fused feature vector; performing dense residual encoding on the initial fused feature vector through an intermediate layer to obtain decoded features; inputting the decoded features into an output layer; and mapping the decoded features to a one-dimensional sea ice concentration space to obtain a sea ice concentration scalar. This method can improve the accuracy of sea ice concentration inversion results.
Need to check novelty before this filing date? Find Prior Art