基于多算法融合的DCT域盲水印FPGA硬件实现
首发时间:2020-05-11
摘要:本文设计了一种相关数据库和具备DCT系数误差校准功能的与改进的线性内插方法相结合的DCT域嵌入算法。实验证明:其对原图破坏较小(嵌入后图像的PSNR值可达30,NC值可达1.0000),并且对于JEPG有损压缩及其他一些恶意攻击都具有良好的鲁棒性。本文在提出一种适于硬件实现的用于图像版权保护的数字盲水印算法的基础上,对其进行了FPGA验证。本设计采用软硬协同的实现方案,同时硬件采用流水线结构,进一步提高了处理速度(1.6倍),非常适合高速信号处理的应用。
关键词: 盲水印 DCT域 数据相关性 线性内插 FPGA硬件实现
For information in English, please click here
FPGA hardware implementation of blind watermarking in DCT domain based on multi-algorithm fusion
Abstract:A DCTdomain\'s embedding algorithm combining relational database and improved linear interpolation with DCT coefficient error calibration was designed.This design can realize an embedded algorithm based on the relational database and the improved linear interpolation Experiments show that it has less damage to the original image (the PSNR value of the embedded image can reach 60, and the NC value can reach 1.0000 ), and it has good robustness against JPEG lossy compression and other malicious attacks.A digital blind watermarking algorithm suitable for hardware implementation for image copyright protection is designed and verified by FPGA.This design adopts a soft-hard cooperative implementation scheme. And the hardware uses a pipeline structure, further improving the processing speed (1.6 times) ,which is very suitable for high-speed signal processing applications.
Keywords: Blind watermarking;DCT domain;Data dependency;Linear Interpolation; Hardware implementation in FPGA
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
基于多算法融合的DCT域盲水印FPGA硬件实现
评论
全部评论