摘要
针对传统快速双循环水平集对初始演化曲线过于依赖的问题,提出一种基于空间惩罚核模糊C-means(SPKFCM)算法的初始演化曲线自动选取快速双循环水平集算法。首先,对模糊均值聚类算法进行改进,通过增加空间惩罚函数提出SPKFCM算法,用于对快速双循环水平集算法的自动初始化;其次,基于SPKFCM并结合快速双循环水平集算法,设计基于SPKFCM快速双循环水平集算法框架,并给出相应速度参量F_d和F_(int)模糊化形式;最后,通过与已有算法在仿真图像上的对比结果显示,所提算法在随机初始化条件下,具有更高的分割精度和计算效率。
In order to solve the problem of the traditional fast two-cycle level set being too much dependent on the initial condition, a kind of space punishment core based C-means algorithm is proposed to automatically select the initial evolution curve for fast two-cycle level set. Firstly, the space penalty core function is used to improve the fuzzy C-means clustering algorithm called the SPKFCM algorithm, which is used for fast two-cycle level set automatic initialization; Secondly, combined SPKFCM and fast two- cycle level set algorithm, the SPKFCM based fast two-cycle level set algorithm is designed, and the fuzzy form of the corresponding speed parametersandare also presented; Finally, it is verified that the proposed algorithm has the higher degree of segmentation precision and higher computation efficiency through simulation with existing contrast algorithms.
出处
《电视技术》
北大核心
2015年第13期27-31,共5页
Video Engineering
基金
河南省教育厅科学技术研究重点项目(14B510028)
河南省科技厅重点科技攻关项目(132102210181)
