段汕,刘丹,周奇鱼.适用于图像恢复的最优形态滤波设计方法[J].中南民族大学学报自然科学版,2019,(2):298-303
适用于图像恢复的最优形态滤波设计方法
Design method of optimal morphological filtering for image restoration
  
DOI:10.12130/znmdzk.20190226
中文关键词: 图像恢复  SV结构映射  SV形态滤波  形态模式谱
英文关键词: image restoration  SV structure mapping  SV morphological filtering  morphological pattern spectrum
基金项目:国家自然科学基金资助项目(61771021)
作者单位
段汕,刘丹,周奇鱼 中南民族大学 数学与统计学学院武汉430074 
摘要点击次数: 137
全文下载次数: 115
中文摘要:
      针对噪声图像的恢复问题,研究了基于SV形态学理论的最优形态滤波设计方法. 通过建立SV形态模式谱解决图像与相关几何结构相似性度量问题;通过引入集差距离函数,解决由形态滤波所产生的几何和拓扑畸变程度的量化问题;通过优化问题的设计,在有效去噪和保持图像基本结构之间寻求一个平衡策略,使得优化问题得以求解. 最终从理论上证明了一类SV交替滤波和一类SV交替惯序滤波是SV形态学方法下,适用于图像恢复的最优滤波.
英文摘要:
      Aiming at the problem of noise image restoration, this paper studies the optimal morphological filtering design method based on SV morphology theory.The similarity measurement problem between image and related geometric structure is solved by establishing SV morphological pattern spectrum, and the quantization problem of geometric and topological distortion caused by morphological filtering is solved by introducing the distance function of set difference. Through the design of the optimization problem, a balance strategy is sought between the effective denoising and preserving the basic structure of the image, and the optimization problem can be solved.Finally, it is proved theoretically that a class of SV alternating filtering and a kind of SV alternating sequence filtering are the optimal filtering for image restoration under the SV morphological method.
查看全文   查看/发表评论  下载PDF阅读器
关闭