基于改进脉冲噪声检测的图像滤波算法

On Image Filtering Algorithm Based on Improved Impulse Noise Detection

在图像降噪处理过程中,针对传统的脉冲噪声检测方法易造成细节模糊和视觉失真等问题,提出一种基于图像分块迭代处理的脉冲噪声检测的滤波方法.该方法对图像局部信息和特征相似性进行相关研究和分析,确定图像局部最优阈值;以局部目标图像为背景,运用图像分割的思想,在不同局部之间选择不同的去噪阈值,根据该阈值从原图像背景中提取图像噪声,进而实现消除噪声的目的.仿真结果表明,与传统的图像去噪方法相比,该方法可以有效地从图像中去除脉冲噪声,同时提取出更多的原始图像信息,很好地保留图像的细节,具有更高的信噪比.

In the process of image noise reduction, for the problem that traditional impulse noise detection method is likely to cause visual distortion and blurred details, image filtering algorithm based on improved impulse noise detection has been presented. Firstly, through the research and analysis of the image partial information and similarity of local image features, local optimum threshold of the image is determined. Then, partial target image as the background, by means of image segmentation, select the type of denois- ing threshold between different locals, the image noise from background of the original image is extracted, thus the purpose of eliminating noise is achieved. The simulation results show that, compared with the traditional image de-noising method, the method can remove impulse noise from the image effectively as well as reserve more details of the original image, and the method has a higher signal noise ratio.