基于分类思想的高密度数据去噪方法研究

DE-NOISE METHOD FOR HIGH DENSITY DATA BASED ON CLASSIFICATION IDEA

这里参照图像去噪方法[1～4],提出了一种基于噪声分类的自适应混合滤波去噪方法。由于传统均值滤波[5]和中值滤波[6]对高斯型噪声和椒盐型噪声(脉冲噪声)有着不同的滤波特性,而在野外采集的原始高密度数据中,可能同时存在高斯型噪声和椒盐型噪声。因此,单独采用中值滤波或均值滤波都不会达到最好的去噪效果。为了能有效滤除这对二种不同性质的噪声,现提出了一种新的混合滤波算法。该算法首先利用局部阈值把受高斯型噪声污染的数据点和受脉冲型噪声污染的数据点区别开来,然后对前者采用均值滤波算法,而对后者则采用带自适应的改进中值滤波算法进行去除。

Referring to the image de-noise methods,the article proposes an adaptive de-noise method with mixed filters based on the noise classification idea.The conventional mean filter and median filter have different filtering characteristics to gauss noise and impulse noise,however,the primitive high-density data from field acquisition may have both gauss noise and impulse noise simultaneously.So,good filtering effect cannot be obtained if we only use mean filter or median filter separately.In order to eliminate these two different types of noise effectively,a new algorithm with mixed filters is proposed.Firstly,The method classify the data points into two classes by local thresholds,one is data set of which the data points are corrupted by gauss noise and the other is data set of which the data points are corrupted by impulse noise.Then,mean filter is used for the former and improved median filter with adaptive traits is used for the latter.