Research On Transform Domain Watermark Algorithm And Hardware Implementation Based On Silicon-Based Laser Chaotic Encryption

The rapid development of the internet has brought convenience to the dissemination of digital products,but it has also caused problems of illegal copying and tampering.Digital watermarking technology is an important technology of copyright protection.It hides specific watermark information in digital products by exploiting the insensitivity of the human visual perception system and the redundancy of digital signals.During disputes,copyright authentication can be performed by extracting watermark information,effectively protecting the rights of the original authors.The two key characteristics of digital watermarking,invisibility and robustness,are mutually constrained.This thesis effectively balances these two characteristics by introducing optimization algorithms.This thesis designs an adaptive non-blind watermarking algorithm and an adaptive blind watermarking algorithm based on the transform domain.Furthermore,it investigates the digital watermark FPGA system based on the discrete wavelet transform and explores the hardware implementation of the digital watermark system.The main research content of this thesis is as follows:(1)An adaptive non-blind watermarking algorithm for color images in the DWT-SVD domain is designed based on silicon-based laser chaotic encryption and improved particle swarm optimization.First,the grayscale watermark image is encrypted using silicon-based laser chaotic sequences and Arnold transformation.Then,the color carrier image is transformed into the YUV color space,and the Y component and the encrypted watermark image are subjected to discrete wavelet transform and singular value decomposition.The singular value matrix of the encrypted watermark is embedded into the singular value matrix of the carrier image using an additive rule.Improved particle swarm optimization is used to find the optimal embedding strength of the watermark,further balancing the invisibility and robustness of the digital watermark.Experimental results show that the designed non-blind watermarking algorithm can meet the requirements of invisibility and can effectively resist image processing attacks such as noise,filtering,JPEG compression,and geometric transformations.(2)A blind watermarking algorithm for color images in the DCT domain is designed based on silicon-based laser chaotic encryption and improved particle swarm optimization.The same encryption method as in the previous section is used for the watermark image,and the mid-frequency coefficients in the DCT domain are selected to embed watermark information.Another chaotic sequences are used to randomly select embedding positions to further enhance the security of the system.Similarly,improved particle swarm optimization is used to calculate the optimal embedding strength of the watermark.Compared with non-blind watermarking,the watermark extraction process of this algorithm does not require the participation of the original image and the watermark image,which improves its practicality.Experimental results show that this algorithm also maintains good invisibility and exhibits good robustness against various image attacks.(3)This thesis explores the hardware implementation of the digital watermark system based on the discrete wavelet transform and proposes an FPGA-based implementation scheme.The key DWT module,Arnold scrambling module,and watermark embedding module were studied and discussed,leading to the FPGA implementation of the watermark system.Both functional simulation and FPGA boardlevel verification indicate the rationality and correctness of the designed watermark system.