ACTIVE FORENSICS METHOD BASED ON LOCAL FEATURES FOR REMOTE SENSING IMAGES

Content and copyright forensics for remote sensing images attract more and more attention. Digital watermarking is the main method of active forensics and copyright protection. In order to improve the robustness and invisibility of image watermarking and avoid modifying original data, this paper proposed a zero watermarking method based on local features. After doing the nonsubsampled contourlet transform, the low frequent sub-band was extracted. This sub-band was divided into several blocks. Discrete cosine transform coefficients of each block were extracted. The direct current component and several low frequent coefficients were used to obtain the feature matrix. The constructed feature matrix and watermark were used to obtain the registration image. In order to improve the security and robustness, original watermark image was scrambled firstly. The experimental results demonstrated the proposed method was robust to common image processing.