摘要
空空导弹红外图像中存在的噪声严重影响目标的检测与识别,论文通过分析灰关联度与图像灰度变化的对应关系,提出了一种基于灰关联分析的红外图像滤波算法。首先选取含噪图像3×3邻域内4个方向的比较序列,分别计算其与理想参考序列之间的灰关联度;再通过比较4个灰关联度的大小得到当前象素点的类型;对噪声点、非噪声点象素分别处理得到去噪后的图像。最后对可见光、红外图像分别添加高斯噪声、椒盐噪声进行去噪实验。结果表明该算法能够很好的去除可见光、红外图像中的高斯噪声及椒盐噪声,具有较强的普适性。
Aim. Different from the rare algorithms that can also filter out Gaussian and salt-and-pepper noises, our novel algorithm is based on grey relational analysis （GRA）. In the full paper, we explain in detail how the GRA effectively removes noises when images are corrupted by Gaussian and salt-and-pepper noises. In this abstract, we just add some pertinent remarks to the two topics of explanation： （1） how the GRA is related to edge detection and how to select sequences according to four operators in 3×3 neighborhood pixels of the image （see Fig 1 in the full paper）; based on this, we calculate the grey relational degrees between the two sequences; （2） through using the grey relational filtering-out algorithm to compare the four directions of grey relational degrees, we distinguish the noise pixels from the non-noise pixels, the processing of which produces images without noise. We also make an experiment on noise removal from a visible light image and an infrared image respectively by adding Gaussian noise and salt- and-pepper noise. The experimental results, given in Figs. 4 and 5 in the full paper, show that our filtering-out algorithm can relatively effectively remove the Gaussian noise and salt-and-pepper noise in both visible light image and infrared image.
出处
《西北工业大学学报》
EI
CAS
CSCD
北大核心
2006年第6期709-712,共4页
Journal of Northwestern Polytechnical University
关键词
红外图像
灰关联分析
高斯噪声
椒盐噪声
灰关联滤波算法
infrared image, grey relational analysis （GRA）, Gaussian noise, salt-and-pepper noise, grey relational filtering-out algorithm
