保持图像细节的局部自适应去噪滤波器新方法

A Local Adaptive Denoising Filter With Keeping Color Images Detail

去噪算法在现代图像处理应用中占有极其重要的地位。然而图像去噪的过程总是伴随着图像的模糊。本文提出了一种对彩色图像细节保持较好的局部自适应的去噪算法--基于个数判断噪声的前提下结合线性插值、非线性插值(这里主要指中值滤波)滤波对噪声图像进行处理,得到较理想的效果。和领域平均法、倒数梯度加权法、纯线性插值法、纯中值滤波法等相比较,其效果改善明显。

Denoising is very important in image processing. Unfortunately denoising will blur the original image, founding a way to keep the details of a noisy image is necessary. We shown a new adaptive algorithm of denoising which has a better performance on keeping image's details-counting the difference of a window combining with the linear interpolation or nonlinear sort filter. We get a better result comparing with mean filter 、 reciprocal gradient filter、linear filter and median filter.