摘要
针对高密度椒盐噪声污染图像的去噪声问题,提出了一种噪声密度估计的梯度检测滤波算法。算法首先对含噪声图像进行总体噪声密度检测,计算噪声密度p,对于低密度噪声图像(p≤40%),采用3×3窗口改进的梯度检测滤波算法对图像进行滤波,对于高密度噪声污染图像,采用5×5窗口改进的梯度检测滤波算法对图像进行滤波。实验结果表明,文中算法对高密度椒盐噪声污染图像具有较强的去噪声能力和细节保持性能,具有较高的实际应用价值。
A filter algorithm based on noise density estimation and gradient detection is proposed to remove high density salt and pepper noise in the image. Firstly,the algorithm detects noise densityp of the noise image,for the low density noise image ( p ≤40% ) ,the improved gradient detection filter algorithm using 3 ×3 windows is applied to filter. For the higher noise image, the algorithm takes 5 ×5 windows to conduct filtering. Experimental results indicate the algorithm has strong denoising ability and good detail maintained performance and has high practical value.
出处
《计算机技术与发展》
2013年第6期82-85,共4页
Computer Technology and Development
基金
博士后基金项目(2012M520158)
辽宁省教育项目(L2012396
L2012397
L2012400)
辽宁省高等学校实验室项目(L2012397)
辽宁省"百千万人才工程"资助项目(2012921058)
关键词
中值滤波
椒盐噪声
噪声密度
梯度
median filter
salt and pepper noise
noise density
gradient