采用自适应四点窗中点滤波的高椒盐噪声滤除方法

发布时间：2018-03-16 03:19

本文选题：图像恢复　切入点：图像去噪　出处：《计算机应用》2017年03期 　论文类型：期刊论文

【摘要】：针对当前中值滤波器处理图像高椒盐噪声效果不佳和实时性不强等问题,提出了一种快速自适应四点窗中点滤波(AFMF)方法。首先,为了降低计算复杂度,使用简单的极值方法检测噪声点;然后,摒弃传统的全点窗口,不用中值滤波,而是在开关滤波和裁剪滤波的基础上,采用新型的非线性滤波方法:中点滤波,简化了算法的流程,提升了算法的计算效率,提高了去噪效果;最后,从3×3窗口开始到由里向外推进,逐渐增大窗口,形成自适应滤波,一直到噪声点处理完,如此避免了窗口大小参数的设置。实验结果表明,与AMF、SAMF、MDBUTMF以及DBCWMF相比,AFMF在处理高密度椒盐噪声上不仅有更好的去噪性能、更快的运行速度(约0.18 s),且无需设置参数,可操作性强,具有较强的实用性。
[Abstract]:In order to reduce the computational complexity, a fast adaptive four-point window mid-point filter (AFMFF) method is proposed to solve the problems of poor performance and poor real-time performance of the median filter in the image processing of high pepper and salt noise. First of all, in order to reduce the computational complexity, a fast adaptive four-point window mid-point filtering method is proposed. Using simple extremum method to detect noise points, then, abandoning the traditional full-point window and using the median filter instead of switching filtering and clipping filtering, a new nonlinear filtering method, midpoint filtering, is adopted. The flow of the algorithm is simplified, the computational efficiency of the algorithm is improved, and the denoising effect is improved. Finally, from the 3 脳 3 window to advancing from the inside to the outside, the window is gradually enlarged to form adaptive filtering until the noise point is processed. The experimental results show that compared with DBCWMF and AMF SAMFMF MDBUTMF, it not only has better denoising performance and faster running speed (about 0.18 sg), but also has strong maneuverability in dealing with high-density salt and pepper noise, the experimental results show that, compared with AMF-SAMFF-MDBUTMF and DBCWMF, it not only has better denoising performance, but also has better operation speed (about 0.18 s-1). It has strong practicability.
【作者单位】：河南师范大学计算机与信息工程学院;河南省高校计算智能与数据挖掘工程技术研究中心;
【基金】：河南省重点科技攻关项目(132102110209) 河南省基础与前沿技术研究计划项目(142300410295)~~
【分类号】：TP391.41;TN713

【相似文献】