Electronics, Год журнала: 2025, Номер 14(9), С. 1853 - 1853
Опубликована: Май 1, 2025
Ensuring the security of remote sensing images is essential to prevent unauthorized access, tampering, and misuse. Deep learning-based digital watermarking offers a promising solution by embedding imperceptible information protect data integrity. This paper proposes SIR-DCGAN, an attention-guided robust method for image protection. It incorporates IR-FFM feature fusion module enhance reuse across different layers SE-AM attention mechanism emphasize critical watermark features. Additionally, noise simulation sub-network introduced improve resistance against common combined attacks. The proposed achieves high imperceptibility robustness while maintaining low computational cost. Extensive experiments on both natural datasets validate its effectiveness, with performance consistently surpassing existing approaches. These results demonstrate practicality reliability SIR-DCGAN secure distribution copyright
Язык: Английский