基于视觉特性的扩展变换量化指纹算法研究  

Research on Spread Transform Quantization Fingerprinting Algorithm Based on the Visual Characteristic

作　　者：程格平[1] 王新颖[1] 杭波[1] 

机构地区：[1]湖北文理学院数学与计算机科学学院,襄阳441053

出　　处：《科学技术与工程》2017年第11期250-254,共5页Science Technology and Engineering

基　　金：国家自然科学基金(61201247);湖北文理学院重点学科开放基金;湖北文理学院教师科研能力培育基金(2016ZK013)资助

摘　　要：针对扩展变换量化指纹算法的性能平衡问题,提出了基于视觉特性的扩展变换量化指纹算法。该算法基于扩展变换的量化指纹算法模型,根据数字图像的局部视觉特性,采用非均匀同步量化的方法调节指纹算法的量化参数,有效减少了指纹图像的量化误差,并实现指纹嵌入检测过程的一致性,从而提高数字指纹的检测效率和载体图像的保真度。实验仿真结果表明,提出的指纹算法明显优于扩展变换量化指纹算法的性能,能够在抗共谋能力和视觉质量之间获得较好的平衡。Aiming at the problems of the performance balance for spread transform quantization fingerprintingalgorithm, a spread transform quantization fingerprinting algorithm based on the visual characteristic is presented. Based on the spread transform quantization fingerprinting model, a non-uniform synchronous quantization method is adopted to adjust the parameters combined with the local visual feather of the digital image, which could effectively reduce the quantization error and realize the consistency of the fingerprinting embedding and detecting process. Accordingly, the detection efficiency of digital fingerprint and the visual fidelity of carrier image are enhanced. The experimental results show that the proposed fingerprinting algorithm is superior to the spread transform quantization fingerprinting algorithm and can obtain a better balance between collusion resistance and perceptual fidelity.

关 键 词：数字指纹 扩展变换 视觉特性 同步量化 算法性能 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]