摘要
针对红外图像噪声复杂多变,在抑制噪声的同时,还需要兼顾细节增强的问题,本文提出了一种基于粗糙集的红外图像多维降噪算法。对采集到的红外图像通过引导滤波进行分层后运用粗糙集理论进一步多维度的分层,分别处理后合并还原得到输出图像。综合对比主观观察与客观评价指标,该算法能够对红外图像降噪有良好效果,对弱小目标细节有良好的增强效果,另外,该算法复杂度较低,具有良好的实时性,在工程实现方面具有良好的应用前景。
With respect to the complexity and variety of infrared image noise,it is necessary to consider the detailed enhancement of images while suppressing noise.Accordingly,this study developed an infrared image denoising algorithm using the rough set theory.Collected infrared images were first layered by guided filtering.Then,further multi-dimensional stratification was conducted using the rough set theory,and output images were obtained through merging and restoration.Compared with a subjective observation and an objective evaluation index,the algorithm was effective in infrared image denoising and helped to enhance weak and small target details.In addition,the algorithm showed low complexity and good real-time performance.It thus has good application prospects in engineering implementations.
作者
王加
周永康
胡健钏
潘超
李泽民
曾邦泽
赵德利
WANG Jia;ZHOU Yongkang;HU Jianchuan;PAN Chao;LI Zemin;ZENG Bangze;ZHAO Deli(Kunming Institute of Physics,Kunming 650223,China)
出处
《红外技术》
CSCD
北大核心
2021年第1期44-50,共7页
Infrared Technology
关键词
红外图像
多维降噪
引导滤波
粗糙集
图像分层
infrared image
multi dimensional noise reduction
guided filtering
rough set
image layering
