基于非局部均值的储粮害虫图像去噪  

Image Denoising of Grain Pest based on Non-local Means

作　　者：黄凌霄[1,2] 张选德[2] 马文涛[2] 

机构地区：[1]宁夏大学土木与水利工程学院,银川750021 [2]宁夏大学数学与计算机学院,银川750021

出　　处：《湖北农业科学》2015年第19期4833-4835,共3页Hubei Agricultural Sciences

基　　金：宁夏大学科学研究基金资助项目(ZR1242);国家自然科学基金项目(61362029;51269024)

摘　　要：介绍了非局部均值(Non-local Means,NLM)去噪算法的基本理论,比较了储粮害虫图像现有的去噪经典方法与非局部均值去噪算法。结果表明,非局部均值去噪算法具有较明显的去噪视觉效果,在峰值信噪比(Peak Signal-to-Noise Ratio,PSNR)和结构相似性(Structural SIMilarity,SSIM)的定量比较中也具有较好的去噪质量。The theoretical basis of non-local means was introduced, and some classical denoising algorithms was compared with the non-local means denoising algorithm in grain insect image. Results showed that non-local means denoising algorithm had the obvious denoising visual effects, also had better denoising quality in the quantitative comparisons of peak signal-tonoise ratio and structural similarity.

关 键 词：非局部均值 图像去噪 峰值信噪比 结构相似性 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]