An Adaptive Watermarking Algorithm Based on Block Kalman Filtering

To solve the security problems such as the illegal duplication, intentional alteration and copyright protection for digital images, this paper proposes a new adaptive paradigm for image watermarking based on the block Kalman filtering. The watermark was disturbed in terms of the logistic rule, and then divided into small blocks. The host image was partitioned into the same small blocks as the ones obtained above. Then, the Kalman filtering model was builded for the watermark blocks and the corresponding host blocks, and used to embed the watermark blocks into the host blocks. The original watermark can be obtained from the embedded hostimage by the inverse Kalman filtering. The proposed algorithm combines the advantages of the Kalman filtering and the benefits of chaos scrambling in order to guarantee the robustness of the watermark. Simulation results show that the proposed method is effective.