基于噪声检测的激光光斑图像去噪方法

Laser spot image filtering method based on noise detection

基于受污染图像的噪声检测,提出了一种有效的椒盐噪声图像混合滤波算法。首先利用可自适应变化的矢量窗口检测噪声,并对检测到的噪声进行分类,然后采用所提出的伪加权中值滤波和伪加权均值滤波两种算法对图像进行混合滤波,最后加入背景阈值和孤立噪点修正量对滤波后的图像进行灰度修正。提出的方法对不同椒盐噪声强度下的激光光斑图像均体现出优异的滤波性能,去噪和边缘保持性能得到了较大提高,优于传统的中值滤波、均值滤波及其一些改进算法。

Based on the contaminated image noise detection, this paper proposed an effective image mixed filtering algorithm for salt and pepper noise. Firstly, the authors used adaptive change vector window detecting the noise and classified the detected noise, then used the pseudo-weighted median filtering and pseudo-weighted mean filtering algorithms for mixed image filtering, finally added background noise threshold and isolated point correct value for image filtering gray correction.The proposed method reflected excellent filtering performance in laser spot images of different salt and pepper noise intensity.The de-noising and edge preserving performance has been greatly improved, better than the traditional median filter, mean filter and some improved algorithms.