Mama-based optically guided synthetic aperture radar image super-resolution reconstruction method and system

Through the Mamba-based optical guidance method, combined with the steps of shallow feature extraction, multimodal feature fusion and high-resolution image reconstruction, the problem of difficulty in making full use of optical image information for SAR image super-resolution reconstruction in the prior art is solved, and higher quality texture detail recovery and performance improvement are achieved.

CN119850422AActive Publication Date: 2025-04-18ANHUI UNIV
View PDF 8 Cites 4 Cited by

Patent Information

Application Number
CN202510030759.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-01-08
Publication Date
2025-04-18
Estimated Expiration
2045-01-08

Smart Images

  • Figure CN119850422A_ABST
    Figure CN119850422A_ABST
Patent Text Reader

Abstract

The invention provides a Mama-based optically guided synthetic aperture radar image super-resolution reconstruction method and system. The method comprises the following steps: a shallow feature extraction step: encoding an SAR image and an optical image into a uniform feature space by using a convolution method; a multi-modal feature fusion step: taking the extracted shallow layer features as input of a CFSM module and an FSAM module, dynamically selecting related and significant optical features, and adaptively performing multi-modal feature fusion in a spatial domain and a frequency domain according to state space parameters so as to generate high-quality fusion features; and a high-resolution image reconstruction step: aggregating the superficial layer features and the fusion features, and then performing up-sampling on the aggregated features so as to reconstruct a high-resolution image. According to the method, the advantage of long-distance modeling of the visual state space equation is utilized, useful information in optical information is selected progressively, cross-modal feature integration is adaptively performed, spatial domain and frequency domain features are effectively utilized, noise interference is reduced, more real texture detail information is recovered, and excellent performance is achieved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Super-resolution method and device, computer equipment and storage medium

    CN113421190A

  • SAR image processing method, device and system and storage medium

    CN114359043A

  • Cross-modal depth forgery detection method based on time-frequency domain visual artifact feature adaptive fusion

    CN114898438A

  • Image super-resolution network training method and device, equipment and storage medium

    CN115601235A

  • 3D magnetic resonance super-resolution method based on cross-modal and cross-scale feature fusion

    CN116029908A

Cited By

  • Photoacoustic imaging artifact removal method and system based on state space and frequency selection

    CN120525743A

  • RGB-T image multi-modal semantic segmentation method based on state space

    CN121305091A

  • Frequency-space combined image fusion method and device and electronic equipment

    CN121353094A

  • Frequency-space joint image fusion method, device and electronic equipment

    CN121353094B