摘要
为了保留数字图像细节特征,且有效抑制目标噪声,提出一种广义混沌映射的数字图像双边全变差滤波算法。通过总结全变差滤波的基本特性,改善双边滤波在阶数邻域中的双重异型滤波融合机制,建立数字双边全变差滤波的最优能量泛函;为了更好的保证数字图像的稳定性,引入规格化参数,重新定义数字图像的距离局部变差,得出适于对广义混沌映射的数字图像的高斯噪声和脉冲噪声两种情形的双边全变差滤波算法。实验表明,所提算法的去噪效果更好,且运行速度较快,可广泛应用与数字图像领域的相关研究。
This paper proposes a digital image bilateral total variation filtering algorithm based on generalized cha-otic map for preserving the detail features of digital image and constraining the target noise effectively.The fusion mechanism of double heteromorphic filtering in order neighborhood was improved to establish the optimal energy func-tional of digital bilateral total variation filtering via summarizing the basic characteristics of total variation filtering.For ensuring the stability of the digital image,the standardization parameter was introduced to redefine the distance local variation of the digital image,thus obtaining the bilateral total variation filtering algorithm for the Gaussian noise and impulse noise of the digital image of the generalized chaotic map.The results show that the algorithm in this work has better denoising effect and faster running speed,which can be widely applied in the digital image field.
作者
何国良
HE Guo-liang(School of Mathematical Sciences UESTC,Chengdu Sichuan 611731,China)
出处
《计算机仿真》
北大核心
2020年第12期245-248,275,共5页
Computer Simulation
关键词
数字图像
边全变差滤波
边缘特征
双重异型滤波机制
最优能量泛函
Digital image
Edge total variation filtering
Edge feature
Double heteromorphic filtering mechanism
Optimal energy functional
