X-band radar echo attenuation correction method based on dual-branch feature enhancement

By constructing a dual-branch feature interaction network and combining it with an edge enhancement mechanism, the attenuation problem of X-band radar under heavy rainfall conditions was solved, achieving high-precision attenuation correction and effective fusion of multi-radar data, thereby improving the reliability and refined representation of radar observations.

CN122307484APending Publication Date: 2026-06-30NANJING UNIV OF INFORMATION SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
NANJING UNIV OF INFORMATION SCI & TECH
Filing Date
2026-05-11
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

X-band radar exhibits significant attenuation under heavy rainfall conditions. Existing attenuation correction methods are insufficient in terms of stability, adaptability, and accuracy, making it difficult to meet the needs of real-time applications. Furthermore, physical consistency is difficult to guarantee when fusing data from multiple radars.

Method used

A deep learning neural network based on bi-branch feature enhancement is adopted. By constructing a bi-branch feature interaction network, multi-scale features of X-band and S-band radar echo data are learned. Attenuation correction is performed by combining edge enhancement mechanism to achieve adaptive modulation and fusion of features.

Benefits of technology

It improves the adaptability to complex precipitation scenarios, enhances the precision and accuracy of attenuation correction results, and strengthens the physical consistency and stability of multi-radar data fusion.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122307484A_ABST
    Figure CN122307484A_ABST
Patent Text Reader

Abstract

This invention discloses an X-band radar echo attenuation correction method based on bi-branch feature enhancement, belonging to the field of meteorological radar data processing technology. The method includes: acquiring X-band and S-band radar echo data from a weather radar and preprocessing the raw radar echo data; constructing an attenuation correction model, inputting the preprocessed X-band radar echo data into the attenuation correction model, and simultaneously introducing corresponding S-band radar echo data as supervisory information to train and test the attenuation correction model to obtain attenuated X-band data; and comparing the attenuated X-band data with the preprocessed S-band radar echo data. This invention effectively alleviates the error problem caused by radar signal attenuation by jointly modeling global and local features and utilizing a cross-branch feature interaction mechanism to enhance information fusion capabilities, significantly improving attenuation correction accuracy while maintaining the integrity of precipitation structure.
Need to check novelty before this filing date? Find Prior Art