段汕,刘丹,周奇鱼.SV形态滤波的参数优化问题[J].中南民族大学学报自然科学版,2021,40(1):102-107
SV形态滤波的参数优化问题
Parameter optimization problem of SV morphological filtering
  
DOI:10.12130/znmdzk.20210116
中文关键词: 图像恢复  SV交替滤波  SV交替惯序滤波  参数优化
英文关键词: image restoration  SV alternating filtering  SV alternating sequential filtering  parameter optimization
基金项目:国家自然科学基金资助项目(61771021)
作者单位
段汕 中南民族大学 数学与统计学学院武汉430074 
刘丹 中南民族大学 数学与统计学学院武汉430074 
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: 103
全文下载次数: 78
中文摘要:
      在SV形态滤波优化问题研究的基础上,进一步研究了SV形态滤波的参数优化问题. 利用图像中噪声颗粒相互隔离的特点和SV形态开、闭的基本性质,通过引入一种新的结构映射序列,缩小了SV形态滤波优化问题中参数的求解范围. 以集差距离函数所确定的恢复误差的上界为优化目标,利用极大极小问题的估计方法,实现对SV形态滤波的上、下界的估计,最终得出了形态滤波的参数优化值.
英文摘要:
      Based on the optimization of SV morphological filtering, the parameter optimization of SV morphological filtering is further studied. Based on the isolation of noise particles in the image and the basic properties of SV shape opening and closing, a new structural mapping sequence is introduced to reduce the range of solution of parameters in SV shape filtering optimization problem. Taking the upper bound of the recovery error determined by the distance function of set difference as the optimization objective, the upper bound and lower bound of SV morphological filtering are estimated by using the estimation method of minimax problem, and finally the optimization value of parameters of morphological filtering is obtained.
查看全文   查看/发表评论  下载PDF阅读器
关闭