摘要
通过图像滤波提高图像的分辨和识别能力,传统的图像滤波算法采用小波降噪方法,由于受到背景色噪声的干扰,小波分解中对低频图像参量的滤波性能不好。提出一种基于Gabor特征分解的高斯混合非线性图像滤波算法。首先进行图像平滑和小波分解预处理,沿梯度方向求得图像边缘信息,在尺度平移平面上进行小波特征分解,得到图像滤波过程中的Gabor小波变换系数,采用高斯混合非线性滤波算法实现图像滤波方法改进。仿真结果表明,采用该方法进行图像滤波,能有效抑制图像斑点噪声,提高图像的分辨性能,对边缘特征和细节的保持能力方面性能有优越,特别适用于对合成孔径雷达成像的滤波处理。
In order to improve the image resolution and recognition ability by image filtering, image filtering algorithm using the traditional wavelet denoising method, due to the interference of background color noise, wavelet decomposition in the fil-tering performance of low-frequency image parameters is not good. This paper puts forward a Gabor feature decomposition nonlinear image filtering algorithm based on Gauss mixture. Firstly, image smoothing preprocessing and wavelet decomposi-tion, obtained along the gradient direction information of image edge, wavelet decomposition characteristics in scale transla-tional plane, Gabor wavelet transform coefficients of image filtering process, using the Gauss hybrid nonlinear filtering algo-rithm and improved image filtering method. The simulation results show that, using the method of image filtering, can effec-tively suppress speckle noise in images, improve image resolution performance, with edge to edge features and details of the ability to maintain performance, especially suitable for synthetic aperture radar imaging processing.
出处
《科技通报》
北大核心
2015年第12期64-66,共3页
Bulletin of Science and Technology
关键词
图像
滤波
小波变换
特征分解
image
filter
wavelet transform
feature decomposition
