<p>Existing frequency-domain watermarking algorithms for remote sensing images exhibit good robustness against conventional image processing operations, yet their resistance to geometric attacks remains insufficient. To address this limitation, this study proposes a novel watermarking algorithm that integrates the strengths of multiple transform domains. The algorithm begins by applying the Harris operator to detect feature points, which are then used to construct feature vectors for geometric correction, thereby enhancing robustness against geometric attacks. Next, the remote sensing image is regarded as three-dimensional data and decomposed along its length, width and band directions using discrete wavelet transform (DWT), followed by discrete cosine transform (DCT) to construct the watermark embedding domain. Finally, singular value decomposition (SVD) is used to further decompose the embedding domain, and the watermark is embedded in the singular values, minimizing energy loss during repeated DCT transformations. Experimental results demonstrate that the watermark is effectively distributed across multiple spectral bands, resulting in excellent invisibility. In terms of robustness, the proposed algorithm exhibits strong resistance to a variety of image processing operations. It can successfully extract the complete watermark under challenging conditions, including JPEG compression with a quality factor of 30, speckle noise with a variance of 0.11, salt-and-pepper noise with a density of 0.11, and gaussian noise with a variance of 0.007. Moreover, the algorithm also maintains high robustness against geometric attacks that alter the image structure, such as rotation, scaling, and cropping.</p>