图像的频域倒数—高斯级联低通滤波去噪方法  

Frequency domain reciprocal—Gaussian cascade low-pass filtering denoising method of image

作　　者：王杰[1] 毛玉泉[1] 李思佳[1] 吴崇虎[1] 

机构地区：[1]空军工程大学信息与导航学院,西安710077

出　　处：《计算机应用研究》2012年第12期4775-4778,共4页Application Research of Computers

基　　金：陕西省电子信息系统综合集成重点实验室基金资助项目(201102Y05);空军工程大学信息与导航学院科研创新基金资助项目(DYCX1040);空军工程大学研究生创新课题资助项目(20110301)

摘　　要：针对图像频域滤波中细节信息丢失的问题,提出了一种频域倒数—高斯级联低通滤波去噪方法。该方法在频域利用倒数快速收敛的性质并结合高斯低通滤波器,实现了图像的联合滤波去噪。其在有效滤除高频噪声的同时更大限度地保留了图像的细节分量,进而使处理后的图像具有较高的对比度,对于去除噪声、提高图像质量有显著的效果。仿真结果表明,在相同的有效滤波面积基础上,与传统低通滤波器、倒数—理想级联低通滤波器、倒数—巴特沃斯级联低通滤波器相比,该方法的去噪效果最佳。To reduce the loss of details in the image frequency domain filtering,this paper proposed a new frequency method that reciprocal—Gaussian cascade low-pass filter.The method utilized the nature of reciprocal fast convergence and combined with Gaussian low-pass filter to achieve the joint filter of the image,it had significant impression on improving image quality by keeping a greater measure of image detail component while filtering the high-frequency noise,it processed the after image ha-ving higher contrast.Simulation shows that,comparing with traditional low-pass filter,reciprocal—ideal cascade low-pass filter and reciprocal—Butterworth cascade low-pass filter based on the same effective filter area,the proposed method is the best on denoising effect.

关 键 词：图像去噪 频域滤波 倒数滤波 有效滤波面积 高斯低通滤波 主观评价法 客观评价法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] TP317.4[自动化与计算机技术—计算机科学与技术]